In vivo transcription on oligomeric simian virus 40 (SV40) DNA

83, 110-119 (1977)

VIROLOGY

In Viva Transcription MOSHE Laboratory

SHANI,

ofBiology

on Oligomeric

MICHAEL

SEIDMAN,

Simian Virus 40 (SV40) DNA AND

NORMAN

of Viruses, National Institute ofAllergy and Infectious Health, Bethesda, Maryland 20014

P. SALZMAN’

Diseases, National

Institutes

of

Accepted June 27,1977 A simian virus 40 nucleoprotein complex containing an endogenous RNA polymerase activity was extracted from infected BSC-1 nuclei with Triton X-100 in low ionic strength buffer. About 30% of both SV40 DNA and total SV40 transcriptional activity in isolated nuclei are recovered by this procedure. When the nucleoprotein complexes are labeled in vitro with [32PlUTP followed by sedimentation in neutral sucrose gradients containing Sarkosyl, labeled RNA is detected as a broad band extending from 21 S to more than 50 S with a major peak at 26 S. After treatment of the labeled complexes with pancreatic ribonuclease, 5 to 8% of the labeled RNA remains associated with the template DNA. The majority of the ribonuclease-resistant material sediments slightly faster than SV40 DNA I. The partially relaxed SV40 DNA I associated with this material represents the predominant template DNA for late SV40 transcription (Birkenmeier et al., 1977). The remainder of the ribonuclease-resistant material sediments in sucrose gradients in a series of faster-sedimenting peaks. The identification of these structures as closed circular oligomeric SV40 DNA molecules that are serving as templates for transcription is based on their sedimentation rates, the time of their accumulation during the lytic cycle, their migration in agarose gels, and their appearance in the electron microscope. At late times after infection at least 10% of the total SV40 template DNA is in the forms of oligomers. The possible correlation between thxe findings and the presence of high molecular weight virus-specific RNA is discussed. INTRODUCTION

Several studies have shown that newly synthesized “late” polyoma- and SV40-specific RNA consists of large heterogenous molecules the bulk of which have molecular weights greater than 1.5 x 106, the size expected for a single transcript of the viral genome (Tonegawa et al., 1970; Acheson et al., 1971; Weinberg et al., 1972; Rozenblatt and Winocour, 1972; Jaenisch, 1972; Rosenthal et al., 1976; Lev and Manor, 1977). This large heterogenous virus-specific RNA is found only in the nucleus (Weinberg et al., 1972; Rozenblatt and Winocour, 1972; Buetti, 1974). The discovery of these late RNA transcripts which are larger than the DNA template raised the possibility that the template for the late transcription might be integrated viral DNA. Some reports have described integration of 1 Author dressed.

to whom reprint

requests

should be ad-

Copyright 0 1977 by Academic press, Inc. All rights of reproduction in any form reserved,

viral DNA into high molecular weight (HMW) cell DNA during lytic infection (Hirai and Defendi, 1972; Ralph and Colter, 1972; Holzel and Sokol, 1974), while other investigations have shown that nonviral sequences are transcribed in tandem with the SV40 DNA (Jaenisch, 1972; Rozenblatt and Winocour, 1972) suggesting the possibility of cotranscription of integrated viral DNA and adjacent cellular DNA similar to that observed in transformed cells (Wall and Darnell, 1971). However, the generation of HMW virusspecific RNA could be accounted for by other mechanisms. Multiple rounds of transcription of viral DNA could occur as a result of an inefficient termination signal similar to in vitro observations with the Escherichia coli RNA polymerase (Fried and Sokol, 1972; Westphal et al., 1973). Alternatively the transcription of tandemly repeated units of SV40 DNA would give rise to HMW viral RNA. 110 ISSN 0042-6822

TRANSCRIPTION

OF OLIGOMERIC

Oligomeric circular DNAs have been previously described in polyoma DNA (Bourgaux, 1973), SV40 DNA (Jaenisch and Levine, 1973; Martin et al., 1976), mitochondrial DNA (Clayton et al., 1970), and bacterial plasmid DNA (Hobom and Hogness, 1974). Martin et al., (1976) have recently shown that during the later stages of productive infection oligomeric forms of SV40 DNA can account for more than 10% of intracellular viral DNA. The exact mechanism of oligomer formation is still unknown although most of the data suggest that. the oligomers arise by intermolecular reciprocal recombination (Hoborn and Hogness, 1974). A major unanswered question regarding these oligomeric forms is their biological role. The present communication demonstrates that the oligomeric supercoiled SV40 DNA molecules can be detected as early as 28 hr postinfection, and that they are utilized as templates for transcription. MATERIALS

AND

METHODS

Tissue cultures and virus. SV40 stocks were prepared using an African green monkey kidney cell line (BSC-1) that was infected wit.h a plaque-purified virus of strain 777 at an input multiplicity of 0.01 plaque-forming units per cell. Cells were grown at 37” in 150-mm plastic petri dishes in Eagle’s medium supplemented with 10% fetal calf serum. Subconfluent cultures were infected with SV40 at an input multiplicity of 5-10 plaque-forming units per cell in Eagle’s medium supplemented with 2% fetal calf serum. Isolation of SV40 transcriptional intermediates. F’orty-eight hours after infection, the cells were trypsinized and washed twice with 0.15 M NaCl, 0.01 M Tris, pH 7.4. Nuclei were prepared as previously described by Shmookler et al. (1974). The nuclei were resuspended volume per volume in 0.04 M Tris, pH 7.4, 0.002 M EDTA, 0.001 M DTT, and 1% Triton X-100. After incubation for 1 hr at 4”, the nuclei were removed by centrifugation at 2400 rpm for 5 min. The supernatant, containing viral nucleoprotein complexes, is referred to as the TET supernatant. RNA synthesis conditions. Assays were

SV40 DNA

111

carried out in a total volume of 125 ~1 as described previously (Shani et al., 1977) with minor modifications. Each reaction mixture contained 46 mM ammonium sulfate; 150 mM KCI; 1.6 mM dithiothreitol; 0.6 mM each of ATP, CTP, and GTP; [(r-32PlUTP (specific activity, 136 to 156 Ci/mmol; New England Nuclear or Amersham/Searle) or [5,6-3HlUTP (specific activity, 41.6 Ci/mmol; New England Nuclear); 56 mM Tris, pH 7.4; and 1.8 mM MnCl, (final concentration was corrected for the amount of EDTA present in the samples). Synthesis occurred for 30 min at 21” and was terminated by adding EDTA to a final concentration of 0.01 M. Sedimentation analysis of the reaction products. After termination of RNA synthesis with EDTA, the reaction mixtures that were not treated with Sarkosyl were layered onto 12-ml linear 5 to 20% (w/w) sucrose gradients containing 0.15 M KCl, 0.02 M Tris, pH 7.4, 1 mM EDTA, 0.5 mM DTT, and 0.5% Triton X-100 and were centrifuged for 2 hr at 40,000 rpm at 4” in a Beckman SW 41 rotor. Tubes were punctured at the bottom and fractions were collected. Aliquots were taken from each fraction and acid-precipitable counts were determined. If the reaction products were treated with Sarkosyl, 25% Sarkosyl was added to the reaction mixture to a final concentration of 0.25%. Then the reaction mixtures were layered onto 12-ml linear 5 to 20% (w/ w) sucrose gradients in 0.05 M Tris, pH 7.4, 0.5 mM DTT, 0.3 M NaCl, and 0.25% Sarkosyl and were centrifuged for 16 hr at 22,000 rpm at 4” in a Beckman SW 41 rotor. Isolation of RNA-DNA hybrid molecules. Reaction products were recovered after sedimentation in sucrose gradients by pooling the indicated fractions and alcohol precipitation. The pellets were dissolved in 200 ~1 of 2x SSC (SSC is 0.15 M NaCI, 0.015 M sodium citrate) and treated with 100 pg/ml of RNase A for 1 hr at 21”. The ribonuclease-treated samples were layered onto 12-ml 5 to 20% (w/w) sucrose gradients containing Sarkosyl, as described above, and were centrifuged for 16 hr at 22,000 rpm at 4” in a Beckman SW 41 rotor.

112

SHANI,

SEIDMAN,

DNA-RNA hybridization. For hybridization experiments SV40 DNA I was prepared as previously described (Lebowitz et al., 1976). DNA was immobilized on 25 mm filters (Millipore HA) which were dried and heated at 80” for 3 hr, and square minifilters (3 x 3 mm) were cut from them. Blank minifilters without DNA were prepared in the same way. The minifilters containing the DNA were incubated with the in vitro-synthesized RNA in a final volume of 100 ~1 of the hybridization buffer (0.1% SDS, 0.01 M Tris, pH 7.4, 2x SSC, and 50% (v/v) formamide) for 72 hr at 37”. At the end of the incubation, the filters were washed with 2x SSC, treated with RNase A (20 pg/ml) at 21” for 1 hr, washed with 2 x SSC, dried, and counted. Agarose-gel analysis of RNA-DNA hybrid molecules. The hybrids were isolated as described above. Peak fractions were pooled and alcohol precipitated. The samples were then electrophoresecl in 0.7% agarose slab gels and autoradiography was performed on vacuum-dried gels as described by Martin et al. (1976). Electron microscopy. DNA was mounted for electron microscopy using the KleinSchmidt aqueous procedure (Davis et al., 1971). Ethidium bromide was added to both spreading and hypophase solutions to a final concentration of 1.5 pg/ml (Sebring et al., 1974). Molecules were traced using a Numonics graphic calculator connected to a Wang 2200 computer system programmed in basic language.

AND

SALZMAN

and nuclei were recovered by pelleting through a sucrose cushion as described (Shmookler et al., 1974). Viral nucleoprotein complexes were released from the nuclei by incubating the nuclei in 0.01 M EDTA, 20 mM Tris, pH 7.4, 0.5 mM clithiothreitol, and 0.5% Triton X-100 (TET) for 1 hr at 4”. The nuclei then were removed by centrifugation at 3400 rpm for 5 min. The TET supernatant contained about 30% of the total viral DNA that was present in infected nuclei. Figure 1 shows the kinetics of the endogenous RNA polymerase activity associated with the TET supernatant. Incorporation of [“HIUTP into acid-precipitable material continued for at least 1 hr at room temperature. Heparin at a concentration of 1 mg/ml stimulated the rate of incorporation of [3H]UTP 1.5-fold or more while the presence of 0.5% Sarkosyl had no effect on the rate of incorporation. The stimulatory effect of heparin is unrelated to its inhibitory effect on nucleases since there was no change in the size of

RESULTS

Preparation and Characterization of SV40 Nucleoprotein Complexes Containing an Endogenous RNA Polymerase Activity Our initial objective was to develop a procedure to isolate SV40 nucleoprotein complexes free of cellular DNA under conditions that minimize release of proteins from the complex. Infected BSC-1 cells were washed at 4” in a hypotonic buffer containing 1 nd4 MgCl*, 25 nd4 Tris, pH 7.4, 0.4 d CaCl,, and 0.5 rd4 dithiothreitol and lysed by Dounce homogenization,

1

15

30 TIME

45

60

(mini

FIG. 1. Rate of incorporation of 13HIUTP into RNA. A TET supernatant was prepared from BSC-1 cells infected with SV40 for 48 hr. A standard RNA polymerase assay was performed as described in Materials and Methods. (---O---O---) TET supernatant; (O- - -0) TET supernatant + 0.5% Sarkosyl; (O--O) TET supernatant + 1.5 mg/ml of heparin.

TRANSCRIPTION

OF OLIGOMERIC

RNA synthesized in the absence of the drug (data not shown). The recovery of SV40 transcriptional activity in the TET supernatant was determined by hybridization to SV40 DNA on filters. Hybridization of [3H]RNA obtained from whole infected nuclei was compared to that of [“HIRNA obtained from TET supernatant in the presence of mock-infected cells. Table 1 shows that about 30% of the total SV40-hybridizable counts was recovered in the TET supernatant. About 75% of the RNA synthesized in vitro in that fraction is SV40 specific (data not shown). These results indicate that a large proportion of the SV40 transcription intermediates can be released from whole nuclei by treatment with Triton and that the maTABLE

1

RECOVERY OF SV40 DNA TRANSCRIPTIONAL

ACTIVITY

AND SV40 IN TET SUPERNATANT

FLUIDS Fraction

PercSe;J;g;itotal Long-labeled DNA

Whole nuclei TET supernatant

100 33

a

Pulse-labeled DNA 100 31

Percentage of 55740 transcriptional activity* 100 21

o BSC-1 cells infected with SV40 were labeled with lL4Clthymidine for 12 hr and with 13H1thymidine for 3.5 min prior to extraction. Nuclei were isolated as described in Materials and Methods. One-half of the nuclei was Hirt extracted; the second half was first extracted with TET buffer and the TET supernatant was treated with SDS. Both fractions were analyzed by neutral and alkali sucrose gradients. The ‘%-labeled DNA contains mostly SV40 DNA I while the 3H-labeled DNA is SV40 replicating molecules. b Nuclei were isolated from infected cells and one-half of the preparation was incubated in the standard RNA polymerase assay for 30 min at room temperature. The second half was used to prepare a TET supernatant. This supernatant was then added to uninfected nuclei. The number of nuclei was equal to that used to prepare the TET supernatant. This mixture was incubated in the standard polymerase assay for 30 min. RNA was extracted by SDSphenol-isoamyl alcohol and hybridized to filters containing SV40 DNA I as described in Materials and Methods.

113

SV40 DNA

jority of the synthesized cific. Sedimentation Analysis SV40 Transcriptional (TI) Complexes

RNA is virus speof

the Active Intermediate

To analyze the sedimentation properties of the active TI, the TET supernatant was first reacted in the standard RNA polymerase reaction with 13HlUTP and then placed on a 5 to 20% (w/w) sucrose gradient containing 0.5% Triton, 20 mM Tris, pH 7.4, 1 mM EDTA, and 0.15 M KC1 and sedimented for 2 hr at 41,000 rpm at 4” in a Beckman SW 41 rotor. Figure 2a shows the size distribution of [3H]UMP-labeled RNA and [‘4Clthymidine-labeled SV40 DNA. The 14C-labeled DNA sediments at a position characteristic of SV40 nucleoprotein complexes, i.e., 50-55 S (Green, 1972). The 3H-labeled RNA sediments more heterogeneously and slightly faster than the labeled viral DNA nucleoprotein complexes. When the reacted TET supernatant is treated with 0.25% Sarkosyl prior to sedimentation and then placed on a 5-20% (w/ w) sucrose gradient containing Sarkosyl (Fig. 2b), the 14C-labeled DNA now cosediments with SV40 DNA I while the 3Hlabeled RNA moves ahead. This profile is very similar to that obtained by the Sarkosyl-extracted SV40 transcriptional intermediates (Shani et al., 1977). Ribonuclease

Treatment

of the TET TIs

Previous results have shown that a fraction of the [3H]UMP-labeled SV40 TI molecules consisting of SV40 DNA I, RNA polymerase II, and nascent RNA chains is resistant to treatment with ribonuclease A. (Shani et aE., 1977; Birkenmeier et al., 1977). The RNase-resistant RNA banded near the 21 S SV40 DNA I and enabled us to identify the major template for SV40 late transcription as covalently closed supercoiled DNA molecules (Birkenmeier et al., 1977). To compare the forms of DNA templates for transcription in the TET supernatant to the DNA template identified by the Sarkosyl procedure (Birkenmeier et al., 19771, labeled TI molecules of the TET supernatant were treated with ribonuclease and analyzed by sucrose

114

SHANI,

SEIDMAN,

AND

SALZMAN

associated with the template DNA. The digestion products were placed on Sarkosyl-sucrose gradients and the profiles are shown in Figs. 3b and c. After treatment of the fast-sedimenting TI molecules with ribonuclease A at least four distinct peaks of C3HlUMP are observed (Fig. 3b) while only one major peak is found with the slower-sedimenting TI molecules (Fig. 3~). The major peak of radioactivity sediments at 22 S and represents the predominant species of SV40 DNA template (Birkenmeier et al., 1977). The three additional peaks of radioactivity sediment at 29, 35, and 39 S. -1

10

20

30

FRACTION NUMBER

FIG. 2. Sedimentation analysis of reaction products. A TET supernatant was prepared from BSC-1 cells infected with SV40 for 48 hr. The cells were prelabeled with [‘Klthymidine for 18 hr prior to extraction. The TET supernatant was incubated in the standard RNA polymerase assay with [5,63H1UTP for 30 min at room temperature in the presence of 1.5 mg/ml of heparin. The reaction was stopped with 0.01 M EDTA; and one-half of the reaction products was sedimented in a sucrose gradient containing Triton X-100 and the other half was sedimented in a sucrose gradient containing Sarkosyl after addition of Sarkosyl to the sample to a final concentration of 0.5%. The gradient containing Triton X-100 (a) was sedimented at 41,000 rpm for 90 min at 4” in a Beckman SW 41 rotor. The gradient containing Sarkosyl (b) was sedimented for 16 hr at 22,000 rpm at 4” in a Beckman SW 41 rotor. (O- - -0) [‘ClThymidine; (0-O) [3HlUMP.

gradient centrifugation. In the experiment outlined in Fig. 3, TET supernatant was labeled in vitro in the standard RNA polymerase reaction and after treatment with 0.25% Sarkosyl was sedimented in a 5-20% Sarkosyl-sucrose gradient (Fig. 3a). Fast-sedimenting TI molecules (30 to 70 S) and slower-sedimenting TI molecules (21 to 30 S) were pooled separately and after alcohol precipitation were subjected to treatment with 100 pg/ml of ribonuclease A for 1 hr in 2~ SSC. Under these conditions, 5 to 8% of the labeled RNA remains

Identification of the Fast-Sedimenting Template Molecules as Oligomeric Forms of SV40 DNA The characterization and identification of the fast-sedimenting template molecules were determined by the time of their appearance during the lytic cycles, from their electrophoretic mobility in agarose gels, and from their appearance in the electron microscope. (a) Time of appearance. It has been shown that oligomeric forms of SV40 DNA accumulate late during the lytic cycle (Martin et al., 1976). To check the time of appearance of the fast-sedimenting peaks, TET supernatants were harvested at different periods after infection, reacted in vitro in the standard RNA polymerase reaction, and, after treatment with 0.25% Sarkosyl, sedimented in Sarkosyl-sucrose gradients. The profile of the TI molecules at different times after infection is shown in Fig. 4a. Incorporation of 13HlUMP increases from 28 hr up to 50 hr after infection and then declines at 62 hr after infection, but there are no differences in the pattern of these profiles. Fast-sedimenting complex molecules from each gradient were pooled as indicated by the arrows and precipitated in alcohol. The pellets were resuspended in 2x SSC and digested with ribonuclease A for 1 hr at room temperature. The digests were then sedimented in Sarkosyl-sucrose gradients. Figure 4b shows the profiles of the ribonuclease-resistant RNA. At 28 hr after infection the majority of the DNA template sediments

TRANSCRIPTION

OF OLIGOMERIC

115

SV40 DNA

20

: CJ -

15

; B 2 3 zt n

10

5

0

/ , b

22s

I

l.61

i7 0 x E 8 -2

12

4 i ?I

, 1

-1

i I A FRACTION

NUMBER

FIG. 3. Sedimentation

analysis of RNA-DNA hybrids. The reaction products that sedimented in sucrose gradients containing Sarkosyl (a) were fractionated into two cuts (I and II) as indicated by the arrows. They were digested with pancreatic ribonuclease A in 2x SSC and sedimented again in sucrose gradients containing Sarkosyl. Cut I (b); and cut II (c). (---O---O---) [“Clthymidine; (m---O) r3H1UMP. Labeling conditions are the same as in Fig. 2.

near the 21 S SV40 DNA I; later in the infection there is a gradual increase in the proportion of faster-sedimenting peaks until 62 hr after infection they comprise about 33% of the total DNA templates in the TET supernatant. (bj Electrophoretic mobility of the ribon&ease-resistant peaks in 0.7% agarose gels. Peak fractions from the ribonuclease-

resistant material were separately collected, concentrated by alcohol precipitation, and subjected to 0.7% agarose slab gel electrophoresis (Fig. 5). As expected, the ribonuclease-resistant peak sedimenting at 22 S migrates similarly to monomeric SV40 DNA template (Birkenmeier et al., 1977) and faster than the other ribonuclease-resistant peaks. The three faster-

116

SHANI,

SEIDMAN,

AND

SALZMAN

nuclease-resistant peaks do not migrate as sharp bands but rather as smears from the position of supercoiled DNA (form I DNA) to the position of the corresponding nicked circular DNA (form II DNA). Migration at this position in the gel is consistent with the idea that the DNA template molecules are partially relaxed. Relaxation of the DNA is due to the attachment of the ribionuclease-resistant RNA to negative supercoiled DNA molecules (Birkenmeier et al., 19771, and partially relaxed supercoiled DNA molecules migrate between form I and the corresponding form II DNA (Keller, 1975). The

AB 6

CD

E

21s +

FRACTION

NUMBER

4. Sedimentation analysis of RNA-DNA hybrids at various times after infection. TET supernatants were prepared at various times after infection and incubated in the standard polymerase reaction with [a-32P11JTP. The reactions were stopped with 0.01 M EDTA, Sarkosyl was added to a final concentration of 0.25%, and the samples were placed on sucrose gradients containing Sarkosyl (a). Fractions 1-19 from each gradient were pooled and alcohol precipitated. The pellets were dissolved in 2x SSC, digested with ribonuclease A, and sedimented again in sucrose gradients containing Sarkosyl (b). (O----O) 26 hr; (O-O) 42 hr; (O- -Cl) 50 hr; (A-.-A) 62 hr.

DNA

II

DNA

I

FIG.

sedimenting peaks migrate in positions characteristic of partially relaxed dimers, trimers, and tetrameric forms of SV40 DNAs (Martin et al., 1976). The four ribo-

FIG. 5. Electrophoresis of RNA-DNA hybrid peaks. Peak fractions from Fig. 4b were separately collected and alcohol precipitated. The pellets were resuspended in 50 ~1 of electrophoresis buffer and subjected to electrophoresis in 0.7% agarose for 10 hr at 60 V as described (Martin et al., 1976). The gel was vacuum dried and autoradiographed with Kodak RP/R2 X-ray film for 2 days. SV40 DNA I and II (A); 22 S peak (B); 29 S peak (0; 35 S peak CD); and 39 S peak (El. Migration is from top to bottom.

TRANSCRIPTION

OF OLIGOMERIC

smearing can be explained by the attachment of different lengths of RNase-resistant RNA to the template DNA. (cl Electron microscopy analysis. Peak fractions of the three different DNA templates were separately collected and after alcohol precipitation, resuspended in 0.01 M Tris, pH 7.4, 0.01 M EDTA. Ethidium bromide (ET-Br) was added to 1.5 pg/ml and the DNA was spread using an aqueous technique. Addition of 1.5 pg/ml of ET-Br to the DNA preparations and to the spreading solution causes the complete relaxation of the supercoiled DNA, which makes it easier to measure contour length (Sebring et al., 1974). The contour length of the DNA taken from the 22 S peak was 1.59 + 0.2 pm, DNA taken from the 29 S peak had a contour length of 3.5 + 0.2 pm, and DNA from the 33 S peak had a contour length of 4.93 + 0.3 pm (Fig. 6). These values are in good agreement with the contour length of monomeric, dimeric, and trimeric SV40 DNA molecules (Martin et al., 1976). DISCUSSION

Many of the circular DNAs form oligomers in which two or more genomes are arranged in a.continuous circle, usually in head-to-tail configuration. Although the question has not been resolved, the forma-

CONTOUR

SV40 DNA

117

tion of oligomers generally has been envisaged in terms of reciprocal recombination between circular DNA molecules (Hobom and Hogness, 1974) or in terms of errors in their replication (Jaenisch and Levine, 1973). However, a major unanswered question was the biological significance of the oligomeric DNA forms inside the cell. The present study clearly demonstrates that oligomeric forms of SV40 DNA serve as a subpopulation of SV40 DNA templates for transcription late in the infection. The predominant DNA template late in the infection is monomeric supercoiled DNA but the proportion of oligomeric DNA template molecules increases with time until at 62 hr after infection a third of the total DNA template molecules in the TET supernatant are oligomers. Since the TET supernatant contains about 30% of the total SV40 transcriptional activity in infected nuclei, then at least 10% of the total transcriptional activity in infected nuclei arises on oligomeric DNA template molecules. Our conclusion that oligomeric SV40 DNA molecules are involved as template molecules for late transcription is based on the assumption that the RNase-resistant RNA that is associated with viral DNA is RNA that is part of the transcription complex. Several lines of evidence have been

LENGTH ,pmI

FIG. 6. Contour-length measurements of DNA taken from the RNA-DNA peaks. Peak fractions from Fig. 4b were separately collected and alcohol precipitated. The pellets were dissolved in 50 ~1 of 0.01 M Tris, 0.01 M EDTA, mounted for electron microscopy, and measured as described in Materials and Methods.

118

SHANI,

SEIDMAN,

presented recently which establish that this is the case and that nonspecific binding of RNA to DNA does not occur (Shani et al., 1977). et al., 1977; Birkenmeier In pulse-chase experiments the sedimentation velocities of the transcription complexes and the size of the labeled RNAs that are bound to them increase during the chase period. However, the RNA that was resistant to RNase at the end of the pulse period becomes progressively more sensitive to RNase during the chase. Measurements of the rate of chain elongation and the finding of a decrease in isotope incorporated into the RNase-resistant RNA during the chase provides clear evidence that the RNase-resistant RNA in TIs represents labeled RNA at the 3’-end of the nascent RNA which is hydrogen bonded to its DNA template. Several lines of evidence have been presented to demonstrate the involvement of oligomeric SV40 DNA molecules in transcription. The sedimentation values of the ribonuclease-resistant peaks are close to those observed for oligomeric SV40 DNAs. The time of appearance of oligomeric SV40 DNA forms coincides with the time of accumulation of these peaks, and the relative mobility of these peaks in 0.7% agarose gels is similar to that of monomeric, dimeric, trimeric, and tetrameric forms of SV40 DNA (Martin et al., 1976). In the electron microscope, the DNA taken from the ribonuclease-resistant peaks is circular with the contour length of the principal peak corresponding to circular SV40 of unit length while the two additional peaks contain closed circles that are, respectively, two and three times that size. In vivo the late polyoma- and SV40-specific RNAs contain high molecular weight RNA (HMW) several times the size of a full transcript of one genome. The origin of this virus-specific HMW-RNA is still unknown. Three models can be considered for the occurrence of this RNA in infected cells. The three models differ in their assumption of the physical state of the viral DNA template. The first model assumes that the HMW virus-specific RNA arises from the cotranscription of integrated viral DNA and adjacent cellular DNA (Jaenisch, 1972; Roz-

AND

SALZMAN

enblatt and Winocour, 1972). Detection of viral DNA sequences cosedimenting with HMW cell DNA in alkaline gradients has been presented as a proof for integration. The demonstration of rapidly sedimenting oligomeric forms of SV40 raises a serious question as to whether rapid sedimentation is adequate evidence either of the occurrence or of the extent of integration during the lytic cycle (Martin et al., 1976). The second model assumes that the HMW virus-specific RNA arises from multiple rounds of transcription of monomeric SV40 DNA. RNA molecules greater than unit length were also detected in vivo with 4X174 DNA (Hayashi and Hayashi, 1970; Clements and Sinsheimer, 19751 and in vitro with $X174 DNA (Smith and Sinsheimer, 1976) and with SV40 DNA (Fried and Sokol, 1972; Westphal et al., 1973). The identification in vivo and in vitro of RNA species with molecular lengths longer than one genome equivalent can be considered as evidence against a totally efficient termination mechanism. The third model assumes that the HMW virus-specific RNA originates from the transcription of tandomly repeated units of SV40 DNA as a result of an inefficient termination mechanism. The data presented here clearly demonstrate the involvement of oligomeric SV40 DNA as a template for late transcription. A definite choice between the last two models cannot be made. A critical experiment which could distinguish between the last two models would be to specifically block the formation of oligomers in infected cells and check for the level of the HMW-RNA. ACKNOWLEDGMENTS We thank Ms. M. Thoren for invaluable technical assistance and Ms. M. Sullivan for her skillful preparation of the electron micrographs. We are grateful to Dr. Jacob Maize1 for helpful discussions. REFERENCES ACHESON, N. H., BUETTI, E., WEIL, R. (1971). Transcription

K., and of the polyoma virus genome: Synthesis and cleavage of giant late polyoma-specific RNA. Proc Nat. Acad. Sci USA 68, 2231-2235. BIRKENMEIER, E. H., RADONOVICH, M. F., SHANI, M., and SALZMAN, N. P. (1977). The SV40 DNA template for transcription of late mRNA in viral SCHERRER,

TRANSCRIPTION

OF OLIGOMERIC

nucleoprotein complexes. Cell 11, 495-504. BOURGAUX, P. ((1973). On the origin of oligomer forms of polyoma DNA. J. Mol. Biol. 77, 197-260. BUETTI, E. (1974). Characterization of late polyoma mRNA. J. Viral. 14, 249-260. CLAYTON, D. A., DAVIS, R. W., and VINOGRAD, J. (1970). Homology and structural relationships between the dimeric and monomeric circular forms of mitochondrial DNA from human leukemic leukocytes. J. Mol. Biol. 47, 137-153. CLEMENTS, J. B., and SINSHEIMER, R. L. (1975). Process of infection with bacteriophage &X174. XXXVII. RN.A metabolism in &X174-infected cells. J. Viral. 15, 151-160. DAVIS, R. W., SIMON, M., and DAVIDSON, N. (1971). Electron microscope heteroduplex methods for mapping regions of base sequence homology in nucleic acids. In “Methods in Enzymology” (L. Grossman and K. Moldave, eds.), Vol. XXI, pp. 413-428. Academic Press, New York. FRIED, A. H., and SOKOL, F. (1972). Synthesis in vitro by bacter:ial RNA polymerase of simian virus IO-specific RNA: Multiple transcription of the DNA template into continuous polynucleotide. J. Gen. Viral. 17, 69-79. GREEN, M. H. (1972). Biosynthetic properties of a polyoma nucleoprotein complex: Evidence for replication sites. J. Virol. 10, 32-41. HAYASHI, Y., and HAYASHI, M. (1970). Fractionation of $X174 specific messenger RNA. Cold Spring Harbor Symp. Qua&. Biol. 35, 171-177. HIRAI, K., and DEFENDI, V. (1972). Integration of simian virus 40 deoxyribonucleic acid into the deoxyribonucleic acid of permissive monkey kidney cells. J. Virol. 9, 705-707. HOBOM, G., and HOGNESS, D. S. (1974). The role of recombination in the formation of circular oligomers of the A dvl plasmid. J. Mol. Biol. 88, 65-87. HOLZEL, F., and SOKOL, F. (1974). Integration of progeny simian virus 40 DNA into the host cell genome. J. Mt?l. Biol. 84, 423-444. JAENISCH, R. (1972). Evidence for SV40-specific RNA containing virus and host-specific sequences Nature New Biol. 235, 46-47. JAENISCH, R., and LEVINE, A. J. (1973). DNA replication of SV40-infected cells. VII. Formation of SV40 catenated and circular dimers. J. Mol. Biol. 73, 199-212. KELLER, W. (1975). Determination of the number of superhelical turns in simian virus DNA by gel electrophoresis. Proc. Nat. Acad. 5%. USA 72, 4876-4880. LEBOWITZ, J.,’ GARON, C. F., CHEN, M. C. Y., and

SV40 DNA

119

SALZMAN, N. P. (1976). Chemical modification of simian virus 40 DNA by reaction with a watersoluble carbodiimide. J. Virol. 18, 205-210. LEV, Z., and MANOR, H. (1977). Amount and distribution of virus-specific sequences in giant RNA molecules isolated from polyoma-infected mouse kidney cells. J. Virol. 21, 831-842. MARTIN, M. A., HOWLEY, P. M. BYRNE, J. C., and GARON, C. F. (1976). Characterization of supercoiled oligomeric SV40 DNA molecules in productively infected cells. Virology 71, 28-40. RALPH, R. K., and COLTER, J. S. (1972). Evidence for the integration of polyoma virus DNA in a lytic system. Virology 48, 49-58. ROSENTHAL, L. J., SALOMON, C., and WEIL, R. (1976). Isolation and characterization of polyAcontaining intranuclear polyoma-specific “giant” RNA. Nucleic Acid. Res. 3, 1167-1183. ROZENBLATT, S., and WINOCOUR, E. (1972). Covalently linked cell and SV40-specific sequences in an RNA from productively infected cells. Virology, 50, 558-566. SEBRING, E. D., GARON, C. F., and SALZMAN, N. P. (1974). Superhelix density of replicating simian virus 40 DNA molecules. J. Mol. Biol. 90,371-379. SHANI, M., BIRKENMEIER, E., MAY, E., and SALZMAN, N. P. (1977). Properties of simian virus 40 transcriptional intermediates isolated from nuclei of permissive cells. J. Virol. 23, 20-28. SHMOOKLER, R. J., Buss, J., and GREEN, M. H. (1974). Properties of the polyoma virus transcription complex obtained from mouse nuclei. Virology 57, 122-127. SMITH, L. H., and SINSHEIMER, R. L. (1976). The in uitro transcription units of bacteriophage 6x174. I. Characterization of synthesis parameters and measurment of transcript molecular weight. J. Mol. Biol. 103, 681-697. TONEGAWA, S., WALTER, G., BERNARDI, A., and, DULBECCO. R. (1970). Transcrintion of the SV40 genome in transformed cells and during lytic infection. Cold Spring Harbor Symp. Quant. Biol. 35, 823-831. WALL, R.. and DARNELL. J. E. (1971). Presence of cell and virus specific sequences in the same molecules of nuclear RNA from virus transformed cells. Nature New Biol. 231, 73-76. WEINBERG, R. A., WARNAAR, S. O., and WINOCOUR, E. (1972). Isolation and characterization of simian virus 40 ribonucleic acid. J. Virol. 10, 193-201. WESTPHAL, H., DELIUS, H., and MULDER, C. (1973). Visualization of SV40 transcription complexes. Lepetit Colloq. Biol. Med. 4, 183-187.