Feb 9, 1977 - Isolated from Nuclei of Permissive Cells. MOSHE SHANI,* EDWARD BIRKENMEIER, EVELYNE MAY, AND NORMAN P. SALZMAN. Laboratory ...
Vol. 23, No. 1 Printed in U.S.A.
JOURNAL OF VIROLOGY, JUly 1977, P. 20-28 Copyright © 1977 American Society for Microbiology
Properties of Simian Virus 40 Transcriptional Intermediates Isolated from Nuclei of Permissive Cells MOSHE SHANI,* EDWARD BIRKENMEIER, EVELYNE MAY, AND NORMAN P. SALZMAN Laboratory ofBiology of Viruses, National Institute ofAllergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20014 Received for publication 9 February 1977
A nucleoprotein complex that is an intermediate in viral transcription has been isolated from simian virus 40 (SV40)-infected BSC-1 cells after lysing infected nuclei with Sarkosyl. It contains DNA, DNA-dependent RNA polymerase II, and nascent RNA chains. RNA chain elongation continues for several hours in vitro and is dependent on exogenous ribonucleoside triphosphates. The complex sediments in neutral sucrose gradients with a main peak at about 24 to 26S. When the nascent RNA on the complex is treated with RNase A, a fraction of the RNA remains resistant to RNase and is hydrogen bonded to the DNA template. The pulse-labeled RNase-resistant RNA can be chased into RNasesensitive RNA, indicating that it is located at the 3' terminus of the RNA chain. The rate of RNA displacement from the DNA template is consistent with an average rate of RNA chain elongation of 15 to 30 nucleotides per min. At least 70% of the RNA synthesized in this in vitro system is SV40 specific. Hybridization with the separated strands of SV40 DNA and with fragments of SV40 DNA generated with endonucleases Hindu + III indicates that this RNA is complementary to all regions of the "late" SV40 DNA strand. Studies of SV40 RNA synthesis in this partially purified preparation at early and late times after infection should provide a way of locating promoter sites for transcription and identifying the form of SV40 DNA that serves as a template for late transcription. RNA transcripts of the "late" strand of simian virus 40 (SV40) DNA are not found in the nucleus and/or the cytoplasm at early times after infection, when a transcript that corresponds to approximately 50% of the "early" strand is detected in both cell fractions (for review, see 1). Late in infection, after DNA synthesis has started, both of the DNA strands are transcribed, but there is 10 to 20 times more nuclear viral RNA complementary to the late strand than to the early strand (2, 6, 9). These data suggest that specific initiation and termination sites exist on the SV40 genome and that transcription of the late strand is more efficient than transcription of the early strand late in infection. Understanding the regulation of the biogenesis of SV40 mRNA requires purification of cell components that can synthesize viral mRNA in vitro. These reconstitution experiments would require the isolation and identification of the RNA polymerases and relevant regulatory proteins in addition to the viral DNA template. However, attempts to create an in vitro system that is a model for in vivo transcription have been unsuccessful. An alternate approach to the problem of studying SV40 RNA synthesis in
vitro is to isolate the in vivo viral complexes of endogenous RNA polymerase, regulatory elements, and the DNA template. After purification of such complexes, one could identify the proteins associated with the template in vivo and determine the physical form of the DNA template. Green and Brooks (7) have extracted a portion of the intracellular SV40 transcription intermediates (TIs). However, most of the viral transcriptional activity remained associated with the cellular DNA and could not be characterized. Recent results of Gariglio and Mousset (5) demonstrated that the majority of SV40 TIs could be extracted from infected nuclei with Sarkosyl. In the present study, we have determined the sedimentation velocity of the SV40 TIs, measured the size of the nascent RNA associated with the TIs before and after treatment with RNase, and hybridized the RNA synthesized in vitro to restriction enzyme fragments of viral DNA. MATERIALS AND METHODS Cells and virus. A line of African green monkey cells (BSC-1) was obtained from M. Singer, and a 20
VOL. 23, 1977
SV40 TRANSCRIPTIONAL INTERMEDIATES
plaque-purified stock of SV40 (small-plaque strain) was obtained from R. G. Martin. Cells were grown at 370C in 150-mm plastic petri dishes in Eagle medium supplemented with 10% fetal calf serum. Infection and labeling of cells. Subconfluent cultures were infected with SV40 at an input multiplicity of 5 to 10 PFU/cell in Eagle medium supplemented with 2% fetal calf serum. The infected cells were labeled with [2-14C]thymidine (1 ACi/ml, 57 mCi/mmol; Radiochemical Centre, Amersham, England) for the time specified. Isolation of SV40 TIs. Forty hours after infection, the cells were trypsinized and washed twice with 0.15 M NaCl and 0.01 M Tris, pH 7.4. Nuclei were prepared by a hypotonic procedure (17) and lysed at 4VC with 0.25% Sarkosyl and 0.4 M NaCl (5). The chromatin was immediately pelleted in an SW50.1 rotor for 30 min at 20,000 rpm at 4VC. The supernatant was carefully poured off and stored at 4VC. Assay for RNA polymerase activity. Assays were carried out in a total volume of 125 A.l that contained: 46 mM ammonium sulfate; 8 mM KCl; 1.8 mM MnCl2; 1.6 mM dithiothreitol; 0.6 mM (each) GTP, CTP, and ATP; 56 iM Tris, pH 7.9; 5 ACi of [5,6-3H]UTP (47 Ci/mmol; New England Nuclear Corp., Boston, Mass.); and 6 ,uM UTP at 22°C. This temperature was used since at higher temperatures there was a rapid loss of enzymatic activity. RNA synthesis was determined by measuring the incorporation of labeled UMP into acid-precipitable material. Purification of RNA synthesized in vitro. At the end of the incubation period, sodium dodecyl sulfate was added to a final concentration of 0.5%, and EDTA was added to 10 mM. The RNA was extracted with phenol-chloroform-isoamyl alcohol (15) and collected by ethanol precipitation. The pellet was dissolved in TKM (0.05 M Tris, pH 7.4; 0.025 M KCl; 0.0025 M MgCl2) and digested with 25 to 50 ,g of DNase per ml (Worthington Biochemicals Corp., Freehold, N.J.; RNase-free and electrophoretically purified) at 2°C for 60 min. The digest was extracted with sodium dodecyl sulfate-phenol and collected by ethanol precipitation. DNA-RNA hybridization. For hybridization experiments with whole viral DNA, SV40 DNA I was prepared as previously described (11). Preparation and purification of SV40 DNA fragments generated by Hind restriction endonuclease was carried out as previously described (13). DNA was immobilized on 25-mm filters (BA85, Schleicher & Schuell Co., Keene, N.H.) that were dried and heated at 80°C for 3 h, and square minifilters (3 by 3 mm) were cut from them. Blank minifilters without DNA were prepared in the same way. The minifilters containing the DNA fragments were preincubated in the hybridization mixture (50% [vol/vol] formamide, 0.3 M NaCl, 0.03 M sodium citrate, 0.1% sodium dodecyl sulfate, and 0.01 M Tris, pH 7.4) for 24 h at 37°C. After this treatment the fragments were irreversibly immobilized on the filters, as judged by the retention of [14Clthymidinelabeled fragments on the filters (13). Hybridization of the in vitro synthesized RNA with either whole SV40 DNA or its Hind fragments was performed, in
a final volume of 100 ,ul of the hybridization buffer, for 72 h at 37°C. At the end of the incubation, the filters were washed with 2x SSC (SSC = 0.15 M NaCl, 0.015 M sodium citrate), treated with pancreatic RNase (20 ,ug/ml) at 22°C for 1 h, washed with 2x SSC, dried, and counted. Sedimentation analysis. Samples were layered onto 11.5-ml linear 5 to 20% (wt/wt) sucrose gradients in 0.3 M NaCl, 0.25% Sarkosyl, 0.001 M dithiothreitol, and 0.01 M Tris, pH 7.4, and centrifuged for 16 h at 24,000 rpm at 4°C in a Spinco SW41 rotor (Beckman Instruments, Inc., Fullerton, Calif.). Fractions were collected, and the amount of acidprecipitable radioactivity in each was determined.
RESULTS Characterization of the transcriptional activity of the Sarkosyl supernatant. BSC-1 cells were infected with plaque-purified SV40, and the Sarkosyl supernatant was prepared 40 h later as described in Materials and Methods. In infected cells that were labeled in vivo for 16 h with [14C]thymidine before the preparation of the Sarkosyl supernatant, more than 85% of the labeled SV40 DNA I was recovered in this supernatant fluid. In the standard assay for endogenous RNA polymerase activity in the Sarkosyl supernatant, incorporation of [3H]UMP into acid-precipitable material continued for at least 2 h at 22°C. Omission of ATP, CTP, and GTP completely inhibited the incorporation of [3H]UMP. Under these conditions, the UTP concentration was rate limiting. The UTP dependency kinetics for the Sarkosyl supernatant RNA polymerase activity is illustrated in Fig. 1. The effect of the concentration of UTP displays classical Michaelis-Menten kinetics. From this data, the Km value for UTP is 19 uM. The presence of a low concentration of aamanitin (0.4 ,Lg/ml), an inhibitor of RNA polymerase II (18), inhibited 96% of the synthetic activity. There was no significant effect caused by high concentrations of rifamycin AF/013 (100 ,ug/ml), an inhibitor that blocks initiation but does not affect RNA chain elongation (14). These results indicate that RNA polymerase II is involved in the transcriptional activity of the Sarkosyl supernatant and that RNA synthesis has been initiated in vivo. These observed data are in agreement with previous reports (5). Sedimentation pattern of the active complex. Forty hours after infection, a Sarkosyl supernatant was prepared and sedimented in a sucrose gradient. Each fraction of the gradient was then assayed for its RNA polymerase activity. Figure 2a shows the sedimentation pattern of the endogenous RNA polymerase activity
SHANI ET AL.
it can be seen that about 60% of the [3H]RNA banded at densities characteristic of RNA-DNA
hybrid structures. Fast-sedimenting complexes
banded at higher densities than slow-sedimenting complexes (data not shown). Free RNA probably is generated during the alcohol precipitation step. Size distribution of. the RNA associated with the active complex. The labeled RNA was extracted from the two pooled fractions (Fig. 2b) with phenol-sodium dodecyl sulfate and treated with DNase. The RNA was then dena-
FIG. 1. UTP kinetics of RNA synthesis of the endogenous RNA polymerase activity in the Sarkosyl supernatant fraction. The effect of UTP concentration on the rate of the RNA polymerase activity was determined in the standard assay mixture, containing 0.6 mMATP, GTP, and CTP. The concentration of UTP was varied from 1 to 10 M. The results are plotted as a double-reciprocal plot according to Lineweaver and Burk.
relative to SV40 DNA I. The RNA polymerase activity is found in several broad peaks, which sediment faster than SV40 DNA I. A main peak of polymerase activity is observed at 24 to 26S. No activity sediments slower than the 21S SV40 DNA I. The biological significance of the two additional peaks of polymerase activity will be presented elsewhere. The sedimentation pattern of the transcriptional complex could also be determined by first incubating the Sarkosyl supernatant with [3H]UTP in the polymerase assay and then sedimenting the reaction products in sucrose gradients. Figure 2b shows labeled RNA sediments as a broad band extending from 21S to more than 40S, with a peak at about 24S. The size heterogeneity of the labeled complex may be due to the attachment of nascent RNA chains of various sizes to the DNA template. Buoyant density analysis of the active complex in CSS04. To investigate whether the transcriptional complex contains nascent RNA chains associated with the DNA template, the buoyant density of the reaction products of an in vitro RNA polymerase reaction was measured in CsS04. After the polymerase reaction, the transcription complexes were first sedimented in sucrose. The peak fractions were pooled and precipitated in 2 volumes of ethanol. The pellet was dissolved in 0.01 M Tris, pH 7.4, and 0.01 M EDTA, and the solution was adjusted to a final density of 1.553 g/cm3 with CsS04 and centrifuged to equilibrium. In Fig. 3
FcIU NUMBER FIG. 2. Sucrose gradient centrifugation analysis of the RNA polymerase activity and the in vitro labeled complex. Forty hours after infection, BSC-1 cells were labeled with [2-'4C]thymidine (1 uCi/ml) (@) for 2 h, and the Sarkosyl supernatant was isolated. (a) The Sarkosyl supernatant was placed on a linear 5 to 20% (wtlwt) sucrose gradient in 0.3 M NaCl, 0.01 M Tris (pH 7.4), 0.25% Sarkosyl, and 0.5 mM dithiothreitol. Centrifugation was for 18 h at 23,000 rpm at 4°C in a Spinco SW41 rotor. Samples of each fraction were assayed for RNA polymerase activity for 30 min at room temperature by incorporation of [5,6-3H]UTP (A) into acid-precipitable material. (b) The Sarkosyl supernatant was incubated in the standard RNA polymerase assay with [5,63HlUTP for 30 min at room temperature. The reaction was stopped with 0.01 M EDTA and placed on a linear sucrose gradient as in (a). The radioactivity in each fraction was determined by trichloroacetic acid
SV40 TRANSCRIPTIONAL INTERMEDIATES
VOL. 23, 1977
20 30 40 FRACTION NUMBER
FIG. 3. Buoyant density of the labeled complex in CsS04. The in vitro labeled complex sedimenting faster than the 21S SV40 DNA marker (Fig. 2a, pool II) was alcohol precipitated. A portion of this material was dissolved in 0.01 M Tris, pH 7.4, and 0.01 M EDTA. CsS04 was added to a final density of1.553 g/cm3, and the sample was centrifuged for 45 h at 40,000 rpm at 15°C in a Spinco SW50.1 rotor. Fractions were collected from the bottom of the tube, and the radioactivity was measured.
tured by heat and formaldehyde treatment and sedimented in sucrose gradients in the presence of formaldehyde (3). The RNA was extremely heterogeneous in size, with a range of 4S to about 20S (Fig. 4). The smaller size of the RNA, compared to in vivo nuclear virus-specific RNA (2), is not generated by RNase activity in the Sarkosyl supernatant, since similar sedimentation properties of the product were obtained in vitro after short and long incubation periods and the sedimentation properties of labeled rRNA incubated together with the Sarkosyl supernatant were not altered. RNA extracted from pool I (Fig. 2b, fastsedimenting complex) sedimented faster than RNA extracted from pool II (slow-sedimenting complex). These results indicate that the size heterogeneity of the labeled complex can be explained, in part, by the attachment of nascent RNA chains of various sizes to the template DNA. RNase sensitivity of RNA in the complex. Free, single-stranded regions of RNA can be converted to acid-soluble fragments by digestion with pancreatic RNase. RNA in stable, double-helical structures is resistant to diges-
tion by this enzyme if the ionic strength is higher than 0.3 M (8). 32P-labeled RNA, associated with the complex after 30 min of incubation in vitro, was pooled (like pool I + II in Fig. 2b) and precipitated in alcohol. The pellet was dissolved in 2 x SSC and incubated with 100 ,ug of pancreatic RNase per ml for 60 min at room temperature. At the end of the digestion the reaction mixture was sedimented in a sucrose gradient. Figure 5a shows the profile of the labeled complex before RNase treatment, and Fig. 5b shows the profile after RNase treatment. A small fraction of the RNA (about 10 to 15%) was resistant to RNase digestion. This fraction sedimented near the 21S SV40 DNA I marker. The size of the RNA resistant to treatment with RNase was determined by polyacrylamide gel electrophoresis. Numerous distinct bands were seen in 12% acrylamide gels, with a size range of 60 to 120 nucleotides and a major band of 90 nucleotides (Birkenmeier, Radonivich, Shani, and Salzman, Cell, in press). Pulse-chase experiments and the maximal rate of RNA elongation in the Sarkosyl supernatant. To demonstrate that the active complex increases in size as a function of the incubation 3 -1.5
15 20 25 FRACTION NLJMRFR
FIG. 4. Size of the RNA synthesized in vitro by the active complex. RNA was extracted from the two pooled fractions (Fig. 2a, pools I [A] and II ). After ethanol precipitation, the RNA was dissolved in 100 p1 of 18% formaldehyde (brought to pH 7.0 with 1 M NaHC03) containing 0.001 M EDTA, heated for 5 min at 700C, and quickly cooled. Before sedimentation, the samples were diluted to 250 p1 with 0.001 M EDTA. The RNA was then fractionated in a 5 to 20% (wt/wt) sucrose gradient in 0.02 M phosphate buffer (pH 7.4), 0.1 M NaCl, and 1% formaldehyde in an SW41 rotor at 32,000 rpm for 19 h at 4°C. The arrows indicate the position of 3Hlabeled 4S and 18S RNA, which were used as sedimentation markers, in a parallel gradient.
SHANI ET AL.
b 8B 6 21 S
FIG. 5. Effect of RNase treatment on the attachment of nascent RNA to the DNA template. (a) The Sarkosyl supernatant was incubated for 15 min at room temperature with [a-32P]UTP in the standard polymerase assay. The reaction was stopped by 0.01 M EDTA, and the reaction mixture was placed on a 36-ml linear 5 to 20% (wt/wt) sucrose gradient and sedimented for 19 h at 23,000 rpm at 40C in a Spinco SW27 rotor. (b) The main peak (fractions 10 to 19) from the gradient shown in (a) was pooled and alcohol precipitated. The pellet was dissolved in 100 /4 of 2 x SSC and treated with 100 pg of pancreatic RNase per ml for 1 h at room temperature. The digest was placed on a 12-ml 5 to 20% (wt/wt) sucrose gradient and centrifuged for 16 h at 26,000 rpm at 4°C in a Spinco SW41 rotor. The arrows indicate the sedimentation values relative to '4C-labeled SV40 DNA I sedimented in a parallel gradient.
time, pulse-chase experiments were performed. If a 40-fold excess of unlabeled UTP was added to the reaction to dilute the [32P]UTP used to start the reaction, no further 3H label was incorporated into RNA (data not shown). The experiment outlined in Fig. 6 shows that a fraction of the radioactivity in the 15-min pulselabeled RNA was in larger complex molecules after 10- and 30-min chases with 2 mM cold UTP. The peak at 24S decreased while there was an increase in the proportion of fastersedimenting complexes. In the previous section we have shown that faster-sedimenting complex molecules contain larger RNA chains than slow-sedimenting complex molecules. Therefore, the increase in the size of the complex under chase conditions is a result of RNA chain elongation.
The rate of RNA chain elongation was determined by following the fraction of the incorporated label that was RNase resistant after various chase periods. After 15 min of synthesis in the presence of [3H]UTP, a sample was taken, and the reaction was stopped with EDTA. To the rest of the reaction mixture cold UTP was added to a final concentration of 2 mM. Samples were removed at various chase periods and sedimented in neutral sucrose gradients. The labeled complexes in these gradients were pooled and ethanol precipitated. The pellets were dissolved in 2 x SSC and digested with 100 lig of pancreatic RNase per ml for 1 h at room temperature. The digestion was stopped with 5% trichloroacetic acid, and the RNase-resistant counts were collected on GF/C filters, washed extensively with 2% trichloroacetic acid, dried, and counted. Figure 7 summarizes the results of two such experiments. The rate of decrease of the RNase-resistant counts during the chase approximates first-order kinetics. After about 2 min, 50% of the RNase-resistant fraction became sensitive to RNase digestion. Only 60% of the RNA that was initially resistant to RNase could be chased, and longer chase periods did not alter that percentage. The reason for this is not clear, but it may reflect inactivation of enzyme molecules on the DNA or dissociation of the enzyme from DNA. 21 S
6 10 EaL
FIG. 6. Sedimentation pattern of the in vitro labeled complex in a pulse-chase experiment. The Sarkosyl supernatant was incubated in the standard RNA polymerase assay with [5,6-3H]UTP for 15 min at room temperature. A sample was taken (zero time) (0), and the reaction was stopped with 0.01 M EDTA. To the remaining reaction mixture cold UTP was added to a final concentration of 2 mM, and samples were taken after 10 (0) and 30 min (A) of incubation. The three samples were placed on linear 5 to 20% (wt/wt) sucrose gradients and sedimented for 18 h at 23,000 rpm at 4°C.
VOL. 23, 1977
SV40 TRANSCRIPTIONAL INTERMEDIATES
these reactions and hybridized with SV40 DNA I immobilized on filters. The results (Table 2) clearly show that both RNA preparations hybridized to the same extent with SV40 DNA I. About 90% of the SV40-hybridizable counts in infected nuclei were recovered in the Sarkosyl supernatant. This indicates that the majority of the viral TIs are solubilized by Sarkosyl extraction. These results are in agreement with those of Gariglio and Mousset (5). The apparent discrepancy between our results and those of Green and Brooks (7) is probably due to different cell lines and differences in extraction
procedures. Hybridization of RNA synthesized by SV40 CHASE TIME IMINJ TIs with Hind fragments of SV40 DNA. The FIG. 7. Kinetics of the decrease of the remaining labeled fraction that is RNase resistant in pulse- regions of the SV40 genome being transcribed 5
chase experiments. The data are from sucrose gradient analyses of two experiments such as shown in Fig. 5 and 6. The radioactive RNA that was RNase resistant at zero time was taken as 100%.
Hybridization of RNA synthesized by the Sarkosyl supernatant with SV40 DNA I. RNA synthesized by the Sarkosyl supernatant is a measure of SV40 transcription. This was suggested by its occurrence in infected cells and not in mock-infected cells (data not shown) and was clearly established by RNA-DNA hybridization experiments. [3H]UMP-labeled reaction products synthesized by the Sarkosyl supernatant were phenol extracted, DNase treated, and sedimented through a sucrose gradient containing 50%o formamide. Labeled RNA sedimenting faster and slower than 18S rRNA marker was pooled separately and hybridized to an excess of SV40 DNA immobilized on filters. Table 1 shows that 69% of the RNA sedimenting slower than 188 rRNA and 25% of the RNA sedimenting faster than 188 rRNA hybridized to SV40 DNA I, compared to 67% hybridization with SV40 complementary RNA. Therefore, the majority of this RNA is virus specific. Green and Brooks (7) demonstrated that most of the virus TIs cannot be solubilized by a variety of extraction procedures, including Sarkosyl. Instead, they remain associated with the cellular DNA in the pellet. In order to compare this with our method, the following experiment was performed. A reaction mixture containing nuclei from infected cells was lysed with 0.25% Sarkosyl, and the entire lysate was incubated in the standard RNA polymerase assay. The second reaction mixture contained a Sarkosyl supernatant obtained from the same number of infected nuclei, which was added to a number of uninfected nuclei equal to that contained in the first reaction mixture, and was assayed under the same conditions. RNA was extracted from
TABLE 1. RNA-DNA hybridization between 3Hlabeled RNA from the RNA polymerase assay and excess amounts of SV40 DNA I Origin of 3H-labeled RNA
Input hybridized cpm % 1,814 25.8 2,740 69.3 67.5 6,915
7,020 >18Sa 18S, representing about 10% of the total RNA synthesized) or slower than 18S (