State of Adenovirus 2 Deoxyribonucleic Acid in theNucleus and Its Mode of Transcription: Studies
with Isolated Viral Deoxyribonucleic Acid-Protein Complexes and Isolated Nuclei
ROBERT D. WALLACE' AND JOSEPH KATES
Department of Chemistry, University of Colorado, Boulder, Colorado 80302
Received for publication 28 December 1971
Newly replicated adenovirus 2 deoxyribonucleic acid (DNA) can be isolatedfrom the nucleus of HeLa cells by a gentle lysis procedure as a fairly homogeneouscomplex with a sedimentation of 73S. The viral DNA complex can be preparedcompletely free from host cell DNA. The viral complex is slightly active in ribo-nucleic acid (RNA) synthesis in vitro. Treatment of the complex with Pronase andsodium dodecyl sulfate converts the DNA to a form which sediments at 43S. Nucleiisolated from adeno-infected cells synthesize high-molecular-weight virus-specificRNA in vitro. Optimal RNA synthesis requires a divalent cation, preferentiallymanganese, and relatively high salt concentrations. The synthesis of virus-specificRNA by the isolated nuclei is strongly inhibited by low doses of a-amanitine.The latter experimental result is discussed in terms of the polymerase used totranscribe the adenovirus DNA in vivo.
Recent studies of viral deoxyribonucleoproteincomplexes isolated from infected cells (viralDNA-protein complexes) have focused attentionon these structures as a means of understandingthe structural and functional organization of theviral genome within the living host cell (3, 4, 9,10, 19). In several of the cases studied, bacteriainfected with T4 (19) and lambda phages (10),or animal cells infected with vaccinia virus (3, 4),the viral DNA-protein complexes which havebeen isolated can support ribonucleic acid (RNA)synthesis in vitro by virtue of an endogenousRNA polymerase activity. This is also true in thecase of a DNA-protein complex isolated frombacteria, consisting principally of the bacterialgenome in a compact state with associated RNApolymerase and nascent RNA chains (14, 21).In general, the RNA synthesized in vitro by allof the above complexes resembles qualitativelythe RNA synthesized in vivo at the time of isola-tion of the DNA-protein complex. In fact, itseems likely that the complexes isolated to dateare only capable of completing RNA chains invitro and not able to initiate new RNA chains.
In this report we describe a procedure for theseparation of an adenovirus DNA-protein com-
plex from the bulk of the cellular chromatin. Theadenovirus DNA-protein complex was isolatedby a modification of the basic procedure used byMelvin Green et al. (9) for the isolation of aDNA-protein complex from polyoma virus-infected cells. The adenovirus complex is freefrom host DNA and is capable of synthesizingsmall quantities of adenovirus RNA in vitro.The second part of this communication de-
scribes the synthesis of adenovirus RNA bynuclei isolated from infected cells. The propertiesof the in vitro reaction are discussed with regardto the question of which RNA polymerase is usedto transcribe the adenovirus DNA in nucleiisolated from infected cells. Although it is likelythat the small nuclear viruses, polyoma andsimian virus 40, must rely on the host cell RNApolymerase, the larger DNA nuclear viruses suchas adenoviruses and herpes viruses may containenough genetic information to code for their ownRNA polymerases. In this study we presentevidence in favor of the utilization of the host cellRNA polymerase II, thought to be located in thenucleoplasm (16), for the transcription, at leastin part, of the adenovirus genome. Similar evi-dence supporting the use of polymerase II foradenovirus transcription has been obtained inde-pendently by Price and Penman (15).
Cells and virus. HeLa S-3 cells were grown inSpinner culture with Eagle F 14 medium (GrandIsland Biological) supplemented with 5%' calf serum.Adenovirus 2, strain adenoid 6, was used throughoutthese studies. The virus was grown in HeLa cells andpurified essentially as described by Green and Pifia(8). Purified virus was stored at 1 X 1017 physicalparticles/ml in a storage buffer containing 0.01 M
tris(hydroxymethyl)aminomethane (Tris)-hydrochlo-ride buffer (pH 8.1), 0.15 M NaCl, 0.1%/ (w/v)bovine serum albumin (crystalline, fraction V, Cal-biochem), and 50%o (v/v) glycerol. The virus was keptat -20 C, but the solution did not freeze. Under theseconditions, no appreciable loss of infectivity was ob-served after a year of storage of a virus preparation.
Cells were infected at a concentration of 1 X107cells/ml after resuspending them in fresh growth me-dium, minus the calf serum. About 25 plaque-formingunits of 500 to 1,000 physical particles of adenovirus 2were added per cell. The adsorption period was 40min at 37 C with stirring. The infected cells were thencentrifuged, and the adsorption medium was dis-carded. The cells were resuspended in growth mediumsupplemented with 5% calf serum at a concentrationof 4 X 105 cells/ml and were incubated at 37 C. Timezero was taken as the time of resuspension of the in-fected cells into the growth medium.
Isolation of nuclei. Two methods were employedfor isolation of nuclei. The first method (Method I),which was used only in some of the intial experiments,was based on breakage of the cells in hypotonic bufferas previously described (11). The second method(Method II) consisted of breaking the cells by gentlepipetting at 5 C for 3 to 5 min in a buffer containing0.01 M Tris-hydrochloride (pH 7.4), 0.6 M sucrose,0.001 M MgCl2, 0.005 M 2-mercaptoethanol, and0.5% (v/v) Triton X-100. The nuclei were collectedby centrifugation at 1,000 X g for 3 min and sus-pended at a concentration of 5 X 107 nuclei/ml in abuffer containing 0.05 M Tris-hydrochloride (pH 7.8),0.004 M 2-mercaptoethanol, and 20%o (v/v) glycerol.Usually, 0.05 ml of this nuclear suspension was usedfor assays ofRNA synthesis in vitro.
Isolation of adenovirus DNA complex. The firstmethod employed consisted of suspension of nuclei at3 X 107/mi in a buffer containing 0.05 M Tris-hydro-chloride (pH 8.0), 0.2 M NaCl, 0.01 M ethylenedia-minetetraacetate (EDTA), and 0.25% Triton X-100(9). Gentle pipetting of the nuclei at 0 C for 3 min wassufficient to cause lysis. The lysed nuclei, about 3 ml,were layered on a 25 to 40% (w/v) sucrose gradientdissolved in the same buffer used for lysis and centri-fuged for 80 min at 10 C in the Beckman SW27 rotorat 25,000 rev/min.
Except for the experiment presented in Fig. 1, allpreparations of adenovirus DNA complexes werecarried out by the following method. Nuclei preparedby the Triton X-100 lysis technique in 0.6 M sucrose(see above) were suspended at 2 X 108/ml in 0.3 M
(NH4)2SO4 containing 0.004 M mercaptoethanol and0.01 M Tris-hydrochloride (pH 7.8). The nuclei wereallowed to stand at 0 C for 3 min and were then di-luted twofold with 0.05 M Tris-hydrochloride (pH 7.8).
After dilution, the partially lysed nuclei were layeredover a cushion of 30% (w/v) sucrose in 0.05 NI Trisand centrifuged for 15 min at 5 C and 20,000 X g.The liquid above the sucrose cushion was collectedwith a Pasteur pipet, and it contained all of the viralcomplex and none of the host cell nuclear DNA. Thisfraction was further fractionated as described in thetext.RNA synthesis by isolated nuclei. The basic reac-
tion mixture contained in a total volume of 0.3 ml:Tris-hydrochloride (pH 7.5), 15 Mmoles; (NH4,)SO4,105 A.moles; MnCl2, 1.4 ,moles; guanosine, cytidine,and adenosine triphosphates (GRP, CTP. ATP), 0.1Mgmole of each; uridine triphosphate (UTP). 0.08,umoles and 2,Ci of 3H:UTP (21 Ci/mmole). About2 X 106 nuclei were present in this reaction mixture,and incubation was for 30 min at 37 C. If higher con-centrations of nuclei were used, clumping and precipi-tation of the nuclei became a serious problem duringthe incubation. The reaction was stopped by the addi-tion of 2 ml of 5% trichloroacetic acid. Tihe acid-in-soluble material was collected on Whatman fiberglassfilters, washed with 20 ml of cold 5% trichloroaceticacid, dried, and counted by liquid scintillatioin spec-troscopy.
Purification of nucleic acids and hybridization ex-periments. RNA was purified from the in vitro reac-tion mixtures as previously described (4). DNA waspurified by the method described by Doerfler (5).DNA-RNA hybridizations were carried out by themethod of Bolle et al. (2). The reactions were incu-bated for 8 hr at 60 C using 10lg of heat-denaturedDNA in each assay for virus-specific RNA.
Sucrose density gradient fractionation of RNA.RNA synthesized in vitro by isolated nuclei was ana-lyzed by centrifugation through a 15 to 30'%, (w/v)sucrose density gradient in 0.01 M Tris-hydrochloride(pH 7.4) containing 0.01 M EDTA, 0.1 NI NaCl, and0.2% sodium dodecyl sulfate (SDS) (NETS buffer).The gradient was centrifuged for 15.5 hr at 16,000rev/min and 25 C in a SW27 rotor of a Beckman ultra-centrifuge. Before being run, the RNA sample was dis-solved in 90% dimethylsulfoxide and quicklyidilutedwith three volumes of the gradient buffer. The latterstep was introduced to minimize the chances of aggre-gation in the RNA sample.
RESULTS
Separation of adenovirus DNA-protein complexfrom cellular DNA. Two methods were usedinitially for the isolation of the adenovirus DNA-protein complexes. Method I (see Materials andMethods) utilized nuclei prepared in hypotonicbuffer, and these nuclei were lysed by resuspen-sion in a buffer which contained 0.25% TritonX-100, 0.2 M NaCl, 0.01 M EDTA, and 0.05 N1Tris-hydrochloride buffer, pH 8.0. The latterlysing procedure was suggested to us bv MelvinGreen of the University of California (La Jolla).To distinguish between host cell DNA and viralDNA, cells were labeled for 4 to 5 hr prior toinfection with 14C-thymidine. Prior to infection
the cells were washed with fresh medium toremove the '4C-thymidine, and at 15 hr postinfec-tion with adenovirus the cells were labeled with3H-thymidine for 2 hr in the presence of 10- Mfluorodeoxyuridine which enhanced incorpora-tion of radioactive thymidine into viral DNA.The time of labeling of viral DNA with 3H-thymidine was chosen on the basis of prior resultswhich showed that host DNA synthesis is almostcompletely inhibited at 15 hr postinfection, whilevirus DNA synthesis occurs at an appreciablerate (7). The doubly labeled nuclear lysate waslayered over a 25 to 40% (w/v) sucrose densitygradient which was centrifuged for 80 min at10 C at 25,000 rev/min in the SW27 rotor of theSpinco ultracentrifuge. Figure 1 illustrates theresults obtained in the latter experiment, showinga complete separation of the viral DNA complexfrom the cellular DNA. The viral DNA countsare found in a rather homogeneous peak in theupper half of the gradient, whereas all of the hostDNA is present in the pellet.A second method (Method II, see Materials
and Methods) for preparation of nuclei andadenovirus DNA-protein complexes was de-veloped and used for all subsequent experimentspresented in this paper. Method II was preferredbecause nuclei prepared by this procedure weremore active in RNA synthesis in vitro and becauseisolation of the adenovirus DNA complex wassomewhat more convenient. After preparation byMethod II, the viral DNA-protein complex wasfurther purified by sedimentation in a 10 to 30%(w/v) sucrose density gradient. The complexsedimented as a fairly homogeneous componentwith a mean sedimentation coefficient of approxi-mately 74S (Fig. 2). This sedimentation value isconsiderably higher than mature adenovirusDNA (32S). Whole adenovirus virions sedimentto the bottom of the gradients employed to bandthe complex.
Effects of treatment of the DNA complex withPronase and SDS. When the adenovirus DNA-protein complex was treated for 30 min with 1mg of heat-treated Pronase per ml and then with0.2% SDS at 37 C for an additional 30 min, theDNA sedimented in a sucrose gradient with asedimentation of about 42 to 45S which is appre-ciably faster than mature viral DNA (Fig. 3).
Synthesis of RNA in vitro by isolated complex.The ability of the viral DNA complex to synthe-size RNA was tested by assaying each fraction ofa sucrose gradient in which the complex hadsedimented as a band for ability to synthesizeRNA from added nucleoside triphosphates.Figure 4 shows a fair degree of correspondencebetween ability to synthesize RNA and thecounts-per-minute profile of the DNA complex
0 5 10 15 20FRACTION
FIG. 1. Separation of viral DNA complex fromcellular DNA. Cellular DNA was labeled with 14C-thymidine for 5 hr prior to infectionz. At the time of in-fection the "4C-thymidine was removed; at 15 hr post-infection thie cells were labeled for 2 hr witht 3H-thymi-dine, andthe tuclei were prepared at 17 hr postinfectionby the hypotonic buffer technique (11). The nuclei werelysed in a buffer containing 0.05 M Tris (pH 8), 0.01 MEDTA, 0.2 m NaCl, anid 0.25% (w/v) Triton X-100(9). A 2-ml amount ofnuclear lysate was layered over a25 to 40% (w/v) sucrose denisity gradient made up inlysis buffer. The gradient was centrifuged in a BeckmanSW27 rotor for 80 min at 10 C and 25,000 rev/min.Fractions (2 mi) were collected from the bottom of thetube, precipitated with 5% trichloroacetic acid, andcollected on fiberglass filters (Whatman GF/C) forcounting. 14C host cell DNA (0); 3H viral DNA (0).
in the gradient. It should be pointed out that atthe peak fraction 6.8 pmoles of uridine mono-phosphate (UMP) was incorporated into RNA in30 min when 2.5 ,ug of DNA, as DNA complex,was added to the assay mixture. This representsa low level of RNA synthesis associated with thecomplex. It should also be noted that somesynthesis of RNA was observed consistently inareas of the gradient which contained little viralDNA. In view of the poor incorporation of pre-cursors into RNA by the complexes, furtherstudies of viral RNA synthesis in vitro werecarried out with isolated nuclei.
Synthesis of RNA by isolated nuclei. Synthesisof RNA in vitro by nuclei isolated from infectedcells at 15 hr postinfection possesses the charac-teristics of a DNA-dependent RNA polymerasereaction (Table 1). Synthesis requires a divalentcation, is inhibited by treatment of the nucleiwith deoxyribonuclease or actinomycin D, and ispartially dependent on the addition of exogenous
FRACTIONFIG. 2. Sedimentation of adenovirus DNA complex
in a sucrose density gradient. Adenovirus DNA com-
plex preparedfrom nuclei lysed in the presence of0.3 M(NH4)2S04 (see Materials and Methods) was centri-fuged through a 10 to 30% (w/v) sucrose gradient ina Beckman SW27 rotor at 25,000 rev/min at 10 C.Ribosomal monomers from HeLa cells were included inthe gradient as sedimentation markers. Sucrose was
dissolved in 0.05 M Tris-hydrochloride, pH 8.0.
nucleoside triphosphates. The requirement forexogenous ATP is only partial, indicating thatthe isolated nuclei may possess a pool of thissubstrate. The product synthesized appears to beRNA since it is completely sensitive to digestionby pancreatic ribonuclease.The nuclei isolated from infected cells can
utilize either MnCl2 or MgCl2. The optimalMnCl2 concentration is 5 mm, and the optimalMgCl2 concentration is 7.5 mm. The activity withMnCl2 is generally somewhat higher than withMgCl2, but the difference in activity is less thantwofold. The reaction in the presence of MnCl2with both infected and uninfected nuclei is en-
hanced by the presence of rather high concentra-tions of (NH4)2SO4 (Fig. 5). The enhancement ofactivity is more pronounced for adeno-infectednuclei than for uninfected nuclei.
Figure 6 demonstrates the time course ofRNAsynthesis by isolated nuclei from infected anduninfected cells. Although the initial rate ofsynthesis appears roughly comparable in bothtypes of nuclei, the nuclei from infected cellscontinue to synthesize RNA for longer periodsof time. The latter observation may account forthe fact that nuclei from adeno-infected cells
prepared at different times after infection showconsiderably more activity in vitro in a 30-minassay than uninfected nuclei. Figure 7 demon-strates the ability of nuclei isolated from infectedcells to synthesize RNA in vitro as a function oftime after infection. The data are presented rela-tive to the amount of RNA synthesized by nucleiisolated from uninfected cells. A maximum in theability to synthesize RNA in vitro is reached withnuclei isolated at about 15 hr postinfection. Atthis time, nuclei isolated from infected cells are
about 4.5 times more active than uninfected cell
nuclei.The fact that the RNA synthesized in vitro by
nuclei from adeno-infected cells represents a highproportion of RNA coded from the viral genomemay be deduced from the hybridization experi-ment shown in Table 2. More than 30% of theRNA synthesized in vitro by nuclei isolated at
-
l
IC
go
C
U-
I
CD
CD
CD
FRACTIONFIG. 3. Sedimentation of the adenovirus DNA com-
plex after treatment with proteolytic enzyme and so-dium dodecyl sulfate (SDS). Viral DNA complex pre-
pared as described in Materials and Methods was
treated with 0.8 mg of heat-treated Pronase per ml (4)at 37 Cfor 30 min. SDS was added to a final concentra-tion of0.2% (w/v), and incubation was continued at 37C for 30 min. The sample was then centrifuged in a su-crose density gradient, 10 to 30% (w/v) in 0.5 M Tris(pH 8.0), for 8 hr at 27,000 rev/min and 10 C in a
Beckman SW27 rotor. Fifty micrograms of mature,intact adenovirus DNA was used as a sedimentationmarker. Counts per minute in DNA from complex(0); optical density at 260 nm of mature adenovirusDNA (0).
FIG. 4. RNA synthesis by adenovirus DNA complex.The DNA complex was prepared and centrifuged asdescribed in legend ofFig. 2. A portion ofeach fractionfrom the gradient was assayed for ability to synthesizeRNA in vitro. Assay conditions used were similar tothose described for isolated nuclei (see Materials andMethods). 14C-thymidine counts/min representing thecomplex DNA (0); 3H-UMP incorporated into RNA(0).
TABLE 1. RNA synthesis in vitro by nuclei isolatedfrom adenovirus 2-infected HeLa cells
Picomoles Per centReaction mixture of 3H-UMP activity ofReactionmixtureincor- complete
several times after infection hybridizes withadenovirus DNA. This estimate of the fraction ofRNA which is adenovirus-specific is likely to below due to possible inefficiency associated with thehybridization experiment.
=3-
a-2
0 0.1 0.2 0.3 0.4Molar (NH4)2 SO4
FIG. 5. Effect of ammonium sulfate on the synthesisof RNA by isolated nuclei. Nuclei from infected cellswere prepared at 15 hrpostinfection. Assay mixture con-tained in a volume of 0.3 ml: Tris-hydrochloride (pH7.5), 15 umoles; (NH4)2S04, 105 Wnoles; MnCl2, 1.5nmoles; GTP, ATP, CTP, 0.1 JAnoles of each; UTP,
0.08 umoles; and 2 uCi of 3H-UTP. Each assay con-tained 2 X 106 nuclei. The reaction mixture was madeup to the desired salt concentration before addition ofthe nuclei. Nuclei from infected cells (*); nuclei fromuninfected cells (0).
0 10 20 30 40 50 60MINUTES
FIG. 6. Time course of RNA synthesis in vitro byisolated nucleifrom infected and uninfected cells. Nucleifrom infected cells were prepared at 18 hr postinfection.Assay conditions were the same as described in legendof Fig. 5. Nuclei from infected cells (0); nuclei fromuninfected cells (0).
FIG. 7. RNA synthesis by nuclei isolatedfrom adeno-virus-infected cells as a function of time after infection.Assay conditions were the same as described in legendof Fig. 5 and incubation was for 30 min at 37 C. Ac-tivity ofthe infected nuclei is expressed as afactor oftheactivity present in uninfected nuclei.
TABLE 2. DNA-RNA hybridization of RNA madein vitro at different times after infection
In vitro RNA Per cent hybridized to Per centprepared from adenovirus 2 DNAa hybridized to
nuclei Heba DNAa
Uninfected 4.4 (0.6)b 3.8 (3.5)b10 hr PIc 22.5 0.712 hr PI 34.7 (10 jeg of DNA) 0.912 hr PI 34.3 (20 ,g of DNA)16 hr PI 28.8 1.320 hr PI 18.3 1.222 hr PI 26 1.3
a Unless otherwise noted, 10 ,ug of denaturedDNA was used. Values represent average ofduplicate points obtained in same experiment.
b Values obtained in a different experiment.c PI = postinfection.
The size distribution of the RNA synthesizedby nuclei isolated from adeno-infected cells at 15hr postinfection is shown in Fig. 8. It is clear thatsome species of RNA with sedimentation coeffi-cients greater than 28S are synthesized. In anotherexperiment, RNA taken from a similar gradientrepresenting fractions greater than 45S in sizewas found to hybridize with adenovirus DNA toan extent comparable to total RNA, beforesucrose fractionation (unpublished data). Thisindicates that adenospecific RNA sequencessynthesized in vitro occur in very large molecules
resembling nucleoplasmic heterogeneous nuclearRNA (Hn RNA) from uninfected cells.
Effect of a-amanitine on RNA synthesis byisolated nuclei. The drug a-amanitine (6, 20, 24)has been demonstrated to be a powerful inhibitorof the polymerase II present in the nucleoplasm ofmammalian cells (12). Addition of the drug toisolated nuclei has been shown to inhibit thesynthesis of Hn RNA in the case of uninfectedHeLa cells, but not to inhibit the synthesis ofribosomal RNA precursors (25), which are syn-thesized in the nucleolus, presumably by RNApolymerase I. In addition to RNA polymerase IIthere is another enzyme, RNA polymerase III,which is thought to be present in the nucleoplasmand which may also synthesize large Hn RNA(15, 16, 25). RNA polymerase III is not inhibitedby a-amanitine (12). In view of the high specificityof a-amanitine inhibition of polymerase II, wedecided to utilize this drug to determine whetherRNA polymerase II was involved in the tran-
=
N
a-
Cm,
0 5 10 15 20FRACTION
FIG. 8. Size distribution in a sucrose density gradientof RNA synthesized in vitro by nuclei isolated fromadenovirus-infected cells. The nuclei used in the in vitroreaction were isolated at 15 hr postinfection. RNA waspurified as described in Materials and Methods andthen centrifuged through2 a 15 to 30% (w/v) sucrosedensity gradient made in NETS buffer for 15.5 hr at16,000 rev/min and 25 C in a Beckman SW27 rotor.Fractions were collected from the bottom of the tube.
scription of the adenovirus genome in vitro. Asimilar experiment could not be carried out invivo because ca-amanitine is not taken up appre-ciably by living HeLa cells.
It may be seen in Fig. 9 that a-amanitine atvery low concentrations inhibited by 80 and 70%the amount ofRNA synthesized in vitro by nucleiisolated from infected and uninfected cells,respectively. This means that a majority of theRNA synthesized in the presence of Mn2+ andhigh (NH4)2SO4 was inhibited by the drug. Theamount of residual RNA synthesis observed inthe presence of the drug could be accounted forby slight continued transcription by the nucleolarpolymerase and by polymerase III of the nucleo-plasm, which are not sensitive to the drug. Thereduced amount of RNA synthesized in thepresence of ae-amanitine by infected nuclei wastested for its ability to hybridize with adenovirusDNA. As shown in Table 3, there was a lower pro-portion of virus-specific RNA in the a-amanitine--resistant fraction, than in the total RNA synthe-sized by uninhibited nuclei from infected cells.Thus, inhibition by at-amanitine of virus-specificRNA closely resembles inhibition of uninfected-cell nuclear RNA synthesis. This is taken asevidence that polymerase II may be utilized, atleast in part, to transcribe the adenovirus DNAin nuclei prepared from infected cells.
100
80
40
_ 20
0 .02 .06 .10jpg of a - Amanitin
FIG. 9. Effect of ce-amanitine onz RNA synthesis inzvitro by nuclei from infected antd uninzfected cells. Reac-tion2 condition2s were the same as described in Fig. 5. a-Amanitine was added before addition2 of nuclei to reac-tiont mixture. RNA syn2thesis by nuclei from infectedcells (0); and iuiiifected cells (0).
TABLE 3. Hy-bridizationl of RNA made in thepresence of ae-amaniitine with the viral DNA
Incorporation into Per centRNA input RNA
N7uclei in reaction' hybridizedNwith ade-Counts/ Per cent novirus 2min control D-NA
Infected 8, 560 100 14Infected plus a- 1,650 19 8amanitine
a Nuclei were prepared at 15 hr postinfection.Ten micrograms of adenovirus 2 DNA were usedfor the hybridization experiments.
DISCUSSIONNewly replicated adenovirus DNA can be
readily isolated free from host cell DNA, as acomplex with a sedimentation coefficient ofapproximately 73S. Treatment of this complexwith Pronase and SDS reduces the sedimentationof the DNA to about 43S. This is considerablyhigher than 32S, the sedimentation coefficient ofmature adenovirus 2 DNA (5). The 43S sedimen-tation coefficient may result from the fact thatthis DNA is twice as long as the mature viralDNA, or, more likely, that this DNA is stillassociated with some other molecular speciesresulting in a higher particle weight and morecompact structure. Perhaps protein correspond-ing to one or more of the internal proteins foundin the adenovirus virion (18) are present in thiscomplex. In fact, the present DNA complex mayresemble, in part, the DNA-protein structureseen in the electron microscope after disruption ofthe adenovirus virion with SDS (23). It is likely,based on this study, that the 73S complex con-tains DNA-dependent RNA polymerase andnascent RNA chains in addition to any otherstructural proteins which might be present. It isclear that further studies of the constituents ofthis complex must be carried out in order toobtain a better understanding of its compositionand structure. A major problem which must beconfronted is the possibility that the complex isformed as a result of the isolation procedures, byadventitious association of host or viral proteinsto the viral DNA. Careful reconstruction experi-ments with exogenously added viral DNA andkinetic studies of the entry of newly synthesizedviral DNA into the complex may be useful ineliminating the latter possibility.RNA synthesis in vitro by nuclei isolated from
infected cells resembles, in some respects, thesynthesis of RNA in vivo. When nuclei wereisolated at different times after adenovirus infec-tion and tested for ability to synthesize RNA
in vitro, the rates observed possessed a maximumat 15 hr postinfection. This maximal rate ofsynthesis was 4.5 times higher than for uninfectedcell nuclei during the 30-min in vitro incubationperiod. RNA synthesis in vivo in adenovirus-infected cells also shows a maximum in the middleof the virus growth cycle (7). This maximumobserved both in vivo and in vitro may result fromthe availability of viral progeny DNA templatewhich might be the limiting factor in RNAsynthesis. The decline in RNA synthesis observedat later times after infection might be due topackaging of the DNA into virions, or cytopathiceffects leading to degeneration of the nucleartranscription system.A substantial fraction of the virus-specific
RNA molecules labeled in vitro are extremelylarge, resembling viral RNA synthesized in vivoin the nucleoplasm (13). Adenovirus RNA iso-lated from polyribosomes, on the other hand, ismuch smaller (13). It is likely that the largenuclear adenovirus RNA represents transcriptsof large segments of the viral genome and istherefore likely to consist of highly polycistronicRNA. In the present study, the question ofbreakdown of the RNA labeled in vitro and itsrelease from the nuclei was not studied in anydetail. It was noted, however, that little, if any,breakdown of the large species of RNA occurredupon incubation at 37 C for 30 min after thereaction of RNA synthesis had gone to comple-tion (unpublished data). The termination of RNAsynthesis in isolated nuclei after about 30 mincould not be reversed by the addition of freshsubstrates. It might be of interest to investigatethe addition of cytoplasmic or nuclear factorsfor their ability to restore RNA synthesis in thespent nuclei.The fact that the isolated nuclei from infected
cells were most active in the presence of Mn2+and high salt [0.3 M (NH4) 2SO4] indicated that anRNA polymerase activity similar to that observedin the nucleoplasm and different from that ob-served in the nucleolus (23, 25) was used in thetranscription of adenovirus DNA. Roeder andRutter (15, 17) have characterized three RNApolymerase species in the nuclei of mammaliancells. Polymerase I is present in the nucleolus.Polymerase II and III are found in the nucleo-plasm. Polymerase II, the major nucleoplasmicspecies, is completely inhibited by low concentra-tions of the drug a-amanitine (12), whereaspolymerase I and III are not sensitive to the drug.It appears likely from recent studies that polym-erase II is involved in the synthesis of Hn RNAin mammalian nuclei (25) and RNA polymeraseIII is also involved in the synthesis of a minorfraction of the Hn RNA. Our results that adeno-
specific RNA synthesis is strongly inhibited inisolated nuclei by ae-aminitine suggest that at leastthe a-amanitine-sensitive moiety of polymerase II
might be used to transcribe the adenovirus DNAin this system. This in turn suggests that polym-erase II is used to transcribe adenovirus genesin vivo. Alternative explanations for the a-
amanitine sensitivity of adenovirus RNA synthe-sis in vitro might be that adenovirus codes for anew RNA polymerase which also happens to besensitive to the drug, or that the in vitro systemis not transcribed by the same polymerase as usedin vivo. However, it seems plausible that a majorpart of the in vitro synthesis is due to completionof RNA chains initiated in vivo. Thus it isreasoned that an a-amanitine-sensitive polym-erase is actually used in vivo to transcribe theviral genome. If the virus does, in fact, use thehost cell nucleoplasmic RNA polymerases, it ispossible that the virus codes for factors whichalter the specificity of the host enzyme(s) so as totranscribe preferentially the viral DNA orselected portions thereof.
ACKNOWLEDGMENTS
The present study was supported by Public Health Service grantROt A108413 from the National Institute of Allergy and InfectiousDiseases.We are grateful to J. Beeson for his valuable assistance. We
thank Melvin Green of the University of California in San Diegofor his advice concerning the methods of isolation of nuclear-virusDNA complexes.
LITERATURE CITED
1. Anderer, F. A., M. A. Koch, and H. D. Schlurnberger. 1968.Structure of simian virus 40. III. Alkaline degradation of thevirus particle. Virology 34:452-458.
2. Bolle, A., R. H. Epstein, W. Salser, and E. P. Geiduschek.1968. Transcription during bacteriophage T# development:requirements for late messenger synthesis. J. Mol. Biol.33:339-362.
3. Dahl, R, and J. Kates. 1970. Intracellular structures contain-ing vaccinia DNA: isolation and characterization. Virology42:453-462.
4. Dahl, R., and J. Kates. 1970. Synthesis of vaccinia virus"early" and "late" messenger RNA in vitro with nucleo-protein structures isolated from infected cells. Virology42:463-472.
5. Doerfier, W. 1969. Nonproductive infection of baby hamsterkidney cells (BHK21) with adenovirus type 12. Virology38:587-606.
6. Fiume, L., and F. Stirpe. 1966. Decreased RNA content inmouse liver nuclei after intoxication with a-amanitin.Biochim. Biophys. Acta 123:643-645.
7. Ginsberg, H. S., L. S. Bello, and A. Levine. 1967. Control ofbiosynthesis of host imiacromolecules in cells infected withadenovirus 5, p. 547-572. In J. S. Colter and W. Paranchych(ed.), The molecular biology of viruses. Academic PressInc., New York.
8. Green, M., and M. Pina. 1963. Adenovirus multiplication. IV.Isolation, purification and chemical analysis of adenovirus.Virology 20:199-207.
9. Green, M. H., H. I. Miller, and S. Hendler. 1971. Isolation ofa polyoma-nucleoprotein complex from infected mouse-cellcultures. Proc. Nat. Acad. Sci. U.S.A. 68:1032-1036.
10. Hayward, W. S.. and M. H. Green. 1969. Effect of lambdarepressor on the binding of RNA polymerase to DNA.Proc. Nat. Acad. Sci. U.S.A. 64:962-969.
11. Kates, J. R., and B. B. McAuslan. 1967. Relationship betweenprotein synthesis and viral DNA synthesis. J. Virol. 1:1 10-114.
12. Lindell, T. J., F. Weinberg, P. W. Morris, R. G. Roeder, andW. J. Rutter. 1970. Specific inhibition of nuclear RNApolymerase II by alpha-amanitin. Science 170:447-449.
13. Parsons, T. J., J. Gardner, and M. Green. 1971. Biochemicalstudies on adenovirus multiplication. XIX. Resolution oflate viral RNA species in the nucleus and cytoplasm. Proc.Nat. Acad. Sci. U.S.A. 68:557-560.
14. Pettijohn, D. E., K. Clarkson, C. R. Kossman, and 0. G.Stonington. 1970. Synthesis of ribosomal RNA on a protein-DNA complex isolated from bacteria: a comparison ofribosomal RNA synthesis in vitro and in vivo. J. Mol. Biol.52:281-300.
15. Price, R., and S. Penman. 1972. Transcription of the adeno-virus genome by an a-amanitine sensitive ribonucleic acidpolymerase in HeLa cells. J. Virol. 9:621-626.
16. Roeder, R. G., and W. J. Rutter. 1969. Multiple forms ofRNA polymerase in eukaryotic organisms. Nature(London) 224:234-347.
17. Roeder, R. G., and W. J. Rutter. 1970. Specific nucleolar andnucleoplasmic RNA polymerases. Proc. Nat. Acad. Sci.U.S.A. 65:675-682.
18. Roeder, R. G., and W. J. Ruttter. 1970. Multiple ribonucleicacid polymerase and ribonucleic acid synthesis during seaurchin development. Biochemistry 9:2543-2553.
19. Russell, W. C., W. G. Lauer, and P. J. Sanderson. 1968. Inter-nal components of adenovirus. Nature (London) 219:1127-1130.
20. Snyder, L., and E. P. Geiduschek. 1968. In, vitro synthesis ofT4 late messenger RNA. Proc. Nat. Acad. Sci. U.S.A.59:459-466.
21. Stirpe, F., and L. Fiume. 1967. Studies on the pathogenesis ofliver necrosis by a-amanitin [Effect of a-amanitin on ribo-nucleic acid synthesis and on ribonucleic acid polymerase inmouse liver nuclei]. Biochem. J. 105:779-782.
22. Stonington, 0. G., and D. E. Pettijohn. 1971. The foldedgenome of Escherichia coli isolated in a protein-DNA-RNAcomplex. Proc. Nat. Acad. Sci. U.S.A. 68:6-9.
23. Van der Eb, A. J., L. W. Van Kesterin, and E. F. J. VanBruggen. 1969. Structural properties of adenovirus DNA's.Biochim. Biophys. Acta 182:530-541.
24. Widnell, C. C., and J. R. Tato. 1964. Evidence for two DNA-dependent RNA polymerase activities in isolated rat-livernuclei. Biochem. Biophys. Acta 87:531-533.
25. Wieland, T. 1964. Peptides of Amanita phalloides. Pure AppI.Chem. 9:145-157.
26. Zylber, E. A., and S. Penman. 1971. The products of the RNApolymerases in HeLa cell nuclei. Proc. Nat. Acad. Sci.U.S.A. 68:2861-2865.
