Proc. Nati. Acad. Sci. USA Vol. 75, No. 4, pp. 1662-1666, April 1978 Biochemistry
Mapping of adenovirus late promoters with nascent mercurated RNA (affinity chromatography/separated strands/hybridization/primary transcripts)
ROBERTO WEINMANN AND LEONARDO 0. AIELLO The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, Pennsylvania 19104
Communicated by Roy Vagelos, January 3, 1978
ABSTRACT Nascent RNA molecules were labeled in vivo and elongated in vitro by incubation of the isolated nuclei in the presence bf mercurated nucleotides. The RNA molecules initiated and labeled in vivo and elongated in vitro were then selectively purified on a thiopropyl 6-B Sepharose affinity column. The procedure was shown to be free of artifacts since the addition of mercurated nucleotides and the retention on the affinity column is mediated by the endogenous RNA polymerase II (nucleoside triphosphate:RNA nucleotidyltransferase; EC 2.7.7.6), is sensitive to actinomycin D, and is dependent on the presence of all four ribonucleotide triphosphates. This general procedure was applied to the mapping of viral promoters late after adenovirus 2 infection of HeLa cells. RNA purified as described above was hybridized to restriction enzyme fragments attached-to nitroceliulose filters. The 5' ends of the nascent RNA chains are located in coordinates 9.5-17 for a rightward transcript, 0-25 for a leftward transcript, and possibly 60-70 for a second rightwa*d transcript. These locations clearly differ from locations of the eatly promoters and therefore suggest that the transition from early to late functions is controlled at the transcriptional level. Control of transcription in prokaryotes occurs at the level of initiation (1). In evkaryotes, a complex series of modifications, including poly(A) addition (2), processing (3), methylation and capping (3), and splicing (4), mark the transition between primary transcripts and mature mRNAs (5). Each of these steps can be controlled independently; in particular, the nature of the signals-controlling the initiation of the RNA chains that are later processed remains obscure. Human cells infected by adenovirus type 2 (Ad 2) have been used as a model system for the study of RNA metabolism in eukaryotes since many features of viral RNA synthesis appear similar to the analogous cellular. processes (6). In this system the host DNA-dependent RNA polymerase II (nucleoside triphosphate:RNA nucleotidyltransferase; EC 2.7.7.6) or a virus-modified RNA polymerase seems to be responsible for the synthesis of viral mRNA precursors (7-9). A strong initiation site (promoter) has been detected at late times after infection by size analysis of large nuclear RNAs (10) or of short nascent RNA chains (11, 12) synthesized after short in vivo or in vitro pulses and hybridized to viral DNA fragments produced with restriction enzymes. Data on the number and location of late promoters have also been obtained by analyzing the effect of UV inactivation of transcription of RNA hybridizing to different regions of the genome (13) or of the inactivation of translation of mRNAs into specific viral proteins for which the gene order is well known (14). The results obtained by these two methods indicate the presence of one or two late promoters,
respectively: An independent procedure for resolving this discrepancy was The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
used here. Mercurated nucleotides (15) were used to selectively purify, by affinity chromatography, the nascent RNA chains prelabeled in vivo, which were then analyzed by hybridization to viral DNA. MATERIALS AND METHODS Cells were prepared and infected with virus as described (8, 9, 16). Labeling of Nuclei Twelve to 14 hr after infection, cells were concentrated 10-fold and labeled with [3H]uridine (Schwarz/Mann, 30-50 Ci/mmol) or [32P]phosphate (New England Nuclear) in phosphate-free medium for 1-3 hr. Nuclei were prepared as described (16) and spun through a 25% glycerol (10 mM Tris/5 mM MgCl2/0.1 mM EDTA/25 mM thioglycerol) cushion. They were resuspended in the cushion buffer at a concentration of 6 X 108 nuclei per ml and used directly in the reactions. Reactions. Nuclei were incubated as described (8) except that the reaction volume was 1-10 ml and UTP or CTP was substituted with HgUTP (15) or HgCTP (17) at a final concentration of 0.4 mM. Preparation of RNA. The reactions incubated at 250 were terminated by the addition of purified DNase (18) (Worthington, DPFF) to a final concentration of 40 jg/ml for 15 min at 40, then by the addition of one volume of 8 M urea sodium dodecyl sulfate (NaDodSO4) adjusted to 0.5% and proteinase K.(Merck) to 100 Aig/ml. Incubation was continued at room temperature for 30-60 min. Then phenol/chloroform/isoamyl extraction, ether extraction, and ethanol precipitation were performed. The precipitates were resuspended in 10 mM Tris (pH 7.4)/5 mM MgCl2 and treated once more with DNase (10 Mig/ml) before they were run on a Sephadex G-50 fine (1 X 20 cm, Pharmacia) column in the same buffer containing 0.2% NaDodSO4 and 5 mM EDTA. The void volume was collected, adjusted to 0.5 M NaCl/0.2% NaDodSO4/1 mM EDTA/50% formamide, and loaded on a thiopropyl-Sepharose 6B (Pharmacia) column (2 X 5 cm) that had been equilibrated in the same buffer at 10 ml/hr and 50°. After extensive washing, the mercurated RNA was eluted with the same buffer used for loading plus 20 ,ug of Escherichia coli tRNA per ml and 0.2 M mercaptoethanol (15, 19). Restriction enzymes Bgl II (from Bacillus globigii), Kpn I (from Klebsiella pneumoniae), Sal I (from Streptomyces albus), EcoRI (from E. coli), and HindIII (from Hemophilus influenza) were obtained from New England Biolabs (Beverly, MA) and incubated according to the directions provided by the supplier. Hybridizations were to restriction enzyme. fragments of labeled Ad 2 DNA attached to nitrocellulose filters as described (20). Hybridization was performed in 0.6 M NaCl/0.06 M sodium citrate at 680 for 20 hr with Denhardt's (21) solution. Abbreviations: NaDodSO4, sodium dodecyl sulfate; Ad, adenovirus.
1662
Biochemistry:
Weinmann and Aiello
RESULTS Rationale. Fig. 1 outlines briefly the rationale for the approach used in these studies. Most of the material prelabeled in vivo is terminated in vivo. Therefore no mercurated nucleotides can be added to it and it is not retained by the affinity column (Fig. 1A, precursors to mRNA). The RNA species initiated and terminated in vitro in the presence of the mercurated nucleotides are not scored since they do not become labeled (first molecule in Fig. 1B). These are mostly the products of transcription by DNA-dependent RNA polymerase III, since little or no reinitiation by RNA polymerase II seems to occur in isolated nuclei (ref. 22; Fig. 1B). The amount of unlabeled nucleotide triphosphates added to the incubation mixture (0.4-0.6 mM) is enough to dilute the small amounts of radioactive precursor that could still remain in the nuclei, since the reaction is completely dependent on the exogenous addition of nucleotide triphosphates (see Table 1, line 4). Only the RNA species that are still being transcribed are labeled in vivo towards the 5' end and mercurated in vitro towards the 3' end; therefore this species is the only radioactive RNA retained on the affinity column. The label in these molecules is preferentially accumulated towards the 5' end and has a distribution similar to that reported by Dintzis for the labeling of nascent proteins (23). Hybridization to viral DNA restriction enzyme fragments should show a peak of hybridization at the 5' end (Fig. 1C) of the RNA. Analysis of In Vitro Reaction. In a typical experiment, cells were labeled, between 14 and 17 hr after infection, with either [3H]uridine or 32P04 and harvested in the cold; nuclei were prepared as described in Materials and Methods. These nuclei were incubated in vitro in the presence of HgCTP (17). Since the mercurated nucleotides by themselves inhibit RNA polymerase (15), it is a mercaptan-HgCTP complex that is used in the transcription reaction. The level of RNA synthesis is approximately 30% of the level found in nuclei incubated in the presence of unmercurated nucleotides and either 2-mercaptoethanol or dithiothreitol as mercaptans (unpublished observations). The reaction can proceed at a fairly good rate, as monitored by the incorporation of [3H]GTP, and continues for 30-60 min (results not shown). Most of the RNA transcribed is a product of endogenous RNA polymerase II, since 80% of the transcriptional activity is inhibited by a-amanitin concentrations (1 ,ug/ml) that completely suppress this enzyme (8, 24). The RNA extracted from these nuclei was broadly distributed around 20 S when analyzed on fully denaturing CH3HgOH-agarose gels (ref. 25; results not shown). The size of the prelabeled nuclear RNA was not affected by incubation for different lengths of time. Since RNA polymerase II in isolated nuclei is unable to reinitiate new RNA chains (22), we assume that HgCTP and [3H]GTP are incorporated into RNA chains preinitiated in vivo. Affinity Chromatography. When RNAs are labeled in vivo and the chains extended in vitro in the presence of HgCTP or HgUTP, the RNAs are retained by a thiopropyl-Sepharose affinity column (Figs. 2 and 3) (17). This retention was not an aggregation artifact, as shown in some other cases (26-28), for the following reasons: (a) [32P]GMP-HgCMP-RNA prepared under similar conditions is retained between 40 and 60% and selectively eluted with 0.2 M mercaptoethanol, while unmercurated cytoplasmic RNA mixed into the reaction and extracted simultaneously is not retained (Fig. 2). Furthermore, when a mixture of [3H]UMP-HgCMP-RNA transcribed from Ad DNA with E. coli RNA polymerase and cytoplasmic [32PIRNA from infected cells was extracted and purified simultaneously, they did not form artificial aggregates under these conditions (Table 1, lines 6 and 7). (b) RNA prepared under similar conditions
Proc. Natl. Acad. Sci. USA 75 (1978)
1663
Precu rsors A In
vivo
to mRNA
Primary
transcripts k-\
RNA
polymerase
-I
\
Viral DNA-
I
T
Nuclei isolation and incubation in the presence of Hg rNTP
B In isolated nuclei
\ )\\I **aCogI-
~~~~~~T
I
RNA extraction and th iopropyl-Sepharose chromatography RNA not retained
:/
....
RNA retained
...~~~~~~....
.....
Hybridization
C Hytbridization
viral
DNA
Map units
FIG. 1. Schematic representation of the transcriptional events detailed in the text. (A) Events occurring in vivo; (B) events occurring in isolated nuclei; (C) results expected upon hybridization of the RNA retained on the affinity column. I, site of initiation of transcription; T, site of termination of transcription; a, RNA polymerase; A, RNA molecules; .... , mercurated RNA chains.
from nuclei incubated in the presence of unmercurated nucleotides or labeled unmercurated cytoplasmic RNA are not retained (less than 0.1%; see Table 1). (c) RNA prepared from nuclei incubated with mercurated triphosphates in the presence of a-amanitin (1 ,tg/ml), to inhibit RNA polymerase II, or in the presence of actinomycin D or in the absence of the nucleTable 1. Effect of in vitro incubation conditions on RNA binding ability to the thiopropyl-Sepharose column Total cpm % bound
Complete reaction + Actinomycin D (40 ,g/ml) + a-Amanitin (1 ,g/ml) -ATP -CTP Cytoplasmic RNA
[3H]UMP-HgCMP-cRNA Cytoplasmic [32P]RNA
711,440 39,263,840 532,400 68,747,658
239,000,000 248,469 1,801,000
RNAs were prepared and chromatographed on a 2
6.58 0.16 0.34 0.21
Mapping of adenovirus late promoters with nascent mercurated RNA (affinity chromatography/separated strands/hybridization/primary transcripts)
ROBERTO WEINMANN AND LEONARDO 0. AIELLO The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, Pennsylvania 19104
Communicated by Roy Vagelos, January 3, 1978
ABSTRACT Nascent RNA molecules were labeled in vivo and elongated in vitro by incubation of the isolated nuclei in the presence bf mercurated nucleotides. The RNA molecules initiated and labeled in vivo and elongated in vitro were then selectively purified on a thiopropyl 6-B Sepharose affinity column. The procedure was shown to be free of artifacts since the addition of mercurated nucleotides and the retention on the affinity column is mediated by the endogenous RNA polymerase II (nucleoside triphosphate:RNA nucleotidyltransferase; EC 2.7.7.6), is sensitive to actinomycin D, and is dependent on the presence of all four ribonucleotide triphosphates. This general procedure was applied to the mapping of viral promoters late after adenovirus 2 infection of HeLa cells. RNA purified as described above was hybridized to restriction enzyme fragments attached-to nitroceliulose filters. The 5' ends of the nascent RNA chains are located in coordinates 9.5-17 for a rightward transcript, 0-25 for a leftward transcript, and possibly 60-70 for a second rightwa*d transcript. These locations clearly differ from locations of the eatly promoters and therefore suggest that the transition from early to late functions is controlled at the transcriptional level. Control of transcription in prokaryotes occurs at the level of initiation (1). In evkaryotes, a complex series of modifications, including poly(A) addition (2), processing (3), methylation and capping (3), and splicing (4), mark the transition between primary transcripts and mature mRNAs (5). Each of these steps can be controlled independently; in particular, the nature of the signals-controlling the initiation of the RNA chains that are later processed remains obscure. Human cells infected by adenovirus type 2 (Ad 2) have been used as a model system for the study of RNA metabolism in eukaryotes since many features of viral RNA synthesis appear similar to the analogous cellular. processes (6). In this system the host DNA-dependent RNA polymerase II (nucleoside triphosphate:RNA nucleotidyltransferase; EC 2.7.7.6) or a virus-modified RNA polymerase seems to be responsible for the synthesis of viral mRNA precursors (7-9). A strong initiation site (promoter) has been detected at late times after infection by size analysis of large nuclear RNAs (10) or of short nascent RNA chains (11, 12) synthesized after short in vivo or in vitro pulses and hybridized to viral DNA fragments produced with restriction enzymes. Data on the number and location of late promoters have also been obtained by analyzing the effect of UV inactivation of transcription of RNA hybridizing to different regions of the genome (13) or of the inactivation of translation of mRNAs into specific viral proteins for which the gene order is well known (14). The results obtained by these two methods indicate the presence of one or two late promoters,
respectively: An independent procedure for resolving this discrepancy was The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
used here. Mercurated nucleotides (15) were used to selectively purify, by affinity chromatography, the nascent RNA chains prelabeled in vivo, which were then analyzed by hybridization to viral DNA. MATERIALS AND METHODS Cells were prepared and infected with virus as described (8, 9, 16). Labeling of Nuclei Twelve to 14 hr after infection, cells were concentrated 10-fold and labeled with [3H]uridine (Schwarz/Mann, 30-50 Ci/mmol) or [32P]phosphate (New England Nuclear) in phosphate-free medium for 1-3 hr. Nuclei were prepared as described (16) and spun through a 25% glycerol (10 mM Tris/5 mM MgCl2/0.1 mM EDTA/25 mM thioglycerol) cushion. They were resuspended in the cushion buffer at a concentration of 6 X 108 nuclei per ml and used directly in the reactions. Reactions. Nuclei were incubated as described (8) except that the reaction volume was 1-10 ml and UTP or CTP was substituted with HgUTP (15) or HgCTP (17) at a final concentration of 0.4 mM. Preparation of RNA. The reactions incubated at 250 were terminated by the addition of purified DNase (18) (Worthington, DPFF) to a final concentration of 40 jg/ml for 15 min at 40, then by the addition of one volume of 8 M urea sodium dodecyl sulfate (NaDodSO4) adjusted to 0.5% and proteinase K.(Merck) to 100 Aig/ml. Incubation was continued at room temperature for 30-60 min. Then phenol/chloroform/isoamyl extraction, ether extraction, and ethanol precipitation were performed. The precipitates were resuspended in 10 mM Tris (pH 7.4)/5 mM MgCl2 and treated once more with DNase (10 Mig/ml) before they were run on a Sephadex G-50 fine (1 X 20 cm, Pharmacia) column in the same buffer containing 0.2% NaDodSO4 and 5 mM EDTA. The void volume was collected, adjusted to 0.5 M NaCl/0.2% NaDodSO4/1 mM EDTA/50% formamide, and loaded on a thiopropyl-Sepharose 6B (Pharmacia) column (2 X 5 cm) that had been equilibrated in the same buffer at 10 ml/hr and 50°. After extensive washing, the mercurated RNA was eluted with the same buffer used for loading plus 20 ,ug of Escherichia coli tRNA per ml and 0.2 M mercaptoethanol (15, 19). Restriction enzymes Bgl II (from Bacillus globigii), Kpn I (from Klebsiella pneumoniae), Sal I (from Streptomyces albus), EcoRI (from E. coli), and HindIII (from Hemophilus influenza) were obtained from New England Biolabs (Beverly, MA) and incubated according to the directions provided by the supplier. Hybridizations were to restriction enzyme. fragments of labeled Ad 2 DNA attached to nitrocellulose filters as described (20). Hybridization was performed in 0.6 M NaCl/0.06 M sodium citrate at 680 for 20 hr with Denhardt's (21) solution. Abbreviations: NaDodSO4, sodium dodecyl sulfate; Ad, adenovirus.
1662
Biochemistry:
Weinmann and Aiello
RESULTS Rationale. Fig. 1 outlines briefly the rationale for the approach used in these studies. Most of the material prelabeled in vivo is terminated in vivo. Therefore no mercurated nucleotides can be added to it and it is not retained by the affinity column (Fig. 1A, precursors to mRNA). The RNA species initiated and terminated in vitro in the presence of the mercurated nucleotides are not scored since they do not become labeled (first molecule in Fig. 1B). These are mostly the products of transcription by DNA-dependent RNA polymerase III, since little or no reinitiation by RNA polymerase II seems to occur in isolated nuclei (ref. 22; Fig. 1B). The amount of unlabeled nucleotide triphosphates added to the incubation mixture (0.4-0.6 mM) is enough to dilute the small amounts of radioactive precursor that could still remain in the nuclei, since the reaction is completely dependent on the exogenous addition of nucleotide triphosphates (see Table 1, line 4). Only the RNA species that are still being transcribed are labeled in vivo towards the 5' end and mercurated in vitro towards the 3' end; therefore this species is the only radioactive RNA retained on the affinity column. The label in these molecules is preferentially accumulated towards the 5' end and has a distribution similar to that reported by Dintzis for the labeling of nascent proteins (23). Hybridization to viral DNA restriction enzyme fragments should show a peak of hybridization at the 5' end (Fig. 1C) of the RNA. Analysis of In Vitro Reaction. In a typical experiment, cells were labeled, between 14 and 17 hr after infection, with either [3H]uridine or 32P04 and harvested in the cold; nuclei were prepared as described in Materials and Methods. These nuclei were incubated in vitro in the presence of HgCTP (17). Since the mercurated nucleotides by themselves inhibit RNA polymerase (15), it is a mercaptan-HgCTP complex that is used in the transcription reaction. The level of RNA synthesis is approximately 30% of the level found in nuclei incubated in the presence of unmercurated nucleotides and either 2-mercaptoethanol or dithiothreitol as mercaptans (unpublished observations). The reaction can proceed at a fairly good rate, as monitored by the incorporation of [3H]GTP, and continues for 30-60 min (results not shown). Most of the RNA transcribed is a product of endogenous RNA polymerase II, since 80% of the transcriptional activity is inhibited by a-amanitin concentrations (1 ,ug/ml) that completely suppress this enzyme (8, 24). The RNA extracted from these nuclei was broadly distributed around 20 S when analyzed on fully denaturing CH3HgOH-agarose gels (ref. 25; results not shown). The size of the prelabeled nuclear RNA was not affected by incubation for different lengths of time. Since RNA polymerase II in isolated nuclei is unable to reinitiate new RNA chains (22), we assume that HgCTP and [3H]GTP are incorporated into RNA chains preinitiated in vivo. Affinity Chromatography. When RNAs are labeled in vivo and the chains extended in vitro in the presence of HgCTP or HgUTP, the RNAs are retained by a thiopropyl-Sepharose affinity column (Figs. 2 and 3) (17). This retention was not an aggregation artifact, as shown in some other cases (26-28), for the following reasons: (a) [32P]GMP-HgCMP-RNA prepared under similar conditions is retained between 40 and 60% and selectively eluted with 0.2 M mercaptoethanol, while unmercurated cytoplasmic RNA mixed into the reaction and extracted simultaneously is not retained (Fig. 2). Furthermore, when a mixture of [3H]UMP-HgCMP-RNA transcribed from Ad DNA with E. coli RNA polymerase and cytoplasmic [32PIRNA from infected cells was extracted and purified simultaneously, they did not form artificial aggregates under these conditions (Table 1, lines 6 and 7). (b) RNA prepared under similar conditions
Proc. Natl. Acad. Sci. USA 75 (1978)
1663
Precu rsors A In
vivo
to mRNA
Primary
transcripts k-\
RNA
polymerase
-I
\
Viral DNA-
I
T
Nuclei isolation and incubation in the presence of Hg rNTP
B In isolated nuclei
\ )\\I **aCogI-
~~~~~~T
I
RNA extraction and th iopropyl-Sepharose chromatography RNA not retained
:/
....
RNA retained
...~~~~~~....
.....
Hybridization
C Hytbridization
viral
DNA
Map units
FIG. 1. Schematic representation of the transcriptional events detailed in the text. (A) Events occurring in vivo; (B) events occurring in isolated nuclei; (C) results expected upon hybridization of the RNA retained on the affinity column. I, site of initiation of transcription; T, site of termination of transcription; a, RNA polymerase; A, RNA molecules; .... , mercurated RNA chains.
from nuclei incubated in the presence of unmercurated nucleotides or labeled unmercurated cytoplasmic RNA are not retained (less than 0.1%; see Table 1). (c) RNA prepared from nuclei incubated with mercurated triphosphates in the presence of a-amanitin (1 ,tg/ml), to inhibit RNA polymerase II, or in the presence of actinomycin D or in the absence of the nucleTable 1. Effect of in vitro incubation conditions on RNA binding ability to the thiopropyl-Sepharose column Total cpm % bound
Complete reaction + Actinomycin D (40 ,g/ml) + a-Amanitin (1 ,g/ml) -ATP -CTP Cytoplasmic RNA
[3H]UMP-HgCMP-cRNA Cytoplasmic [32P]RNA
711,440 39,263,840 532,400 68,747,658
239,000,000 248,469 1,801,000
RNAs were prepared and chromatographed on a 2
6.58 0.16 0.34 0.21
