JOURNAL OF VIROLOGY, Feb. 1975, p. 305-315 Copyright 0 1975 American Society for Microbiology

Vol. 15, No. 2 Printed in U.S.A.

Vaccinia Virus Infection of HeLa Cells I. Synthesis of Vaccinia DNA in Host Cell Nuclei PAOLO LACOLLA' AND ARTHUR WEISSBACH* Roche Institute of Molecular Biology, Department of Cell Biology, Nutley, New Jersey 07110 Received for publication 17 June 1974

The replication of vaccinia virus is thought to take place exclusively in the cytoplasm of host cells. However, using DNA-DNA hybridization techniques, it can be shown that a significant fraction of the synthesis of vaccinia DNA takes place in the nucleus as well as the cytoplasm. The [3H ]thymidine pulse-labeled vaccinia DNA synthesized in the nucleus reaches a maximum at about 3 h after infection, corresponding to the time of maximal DNA synthesis in infected cells. At this time host DNA synthesis drops to about 25% of the rate of the uninfected cells. Even with short labeling times (2 min) the nucleus is found to contain 60% of the incorporated [3H]thymidine, much of which is in vaccinia DNA. Prior inhibition of host nuclear DNA synthesis with mitomycin C, followed by removal of the antibiotic, causes a subsequent inhibition of vaccinia DNA synthesis and complete suppression of mature virus. Purified nuclei, isolated from vacciniainfected cells, also synthesize vaccinia DNA in vitro. Over 90% of the DNA synthesized in vitro by isolated nuclei contain vaccinia-specific sequences. The examination of vaccinia-infected cells by autoradiography (4) and fluorescent antibody staining (20) indicated that vaccinia DNA replicates and virions undergo maturation in large cytoplasmic aggregates (factories). Subsequently, Joklik and Becker (16) demonstrated that the replicating viral DNA, as well as the parental viral DNA, is associated with large cytoplasmic structures which are presumably the same as the vaccinia virosomes identified by autoradiography and electron microscope studies. It is from these factories, starting at 3 to 4 h postinfection, that progressively increasing amounts of viral DNA are released and become combined with the viral core and coat proteins. Dahl and Kates (6, 7) extended these studies by determining the sedimentation and the density characteristics of parental and newly replicated vaccinia DNA. They showed that the isolated cytoplasmic viral DNA-protein complexes synthesize mRNA sequences with an endogenous RNA polymerase. The mRNA synthesized in vitro is either "early" or "early and late" in accordance with the time of virosome isolation, i.e., before or after vaccinia DNA synthesis has commenced. Moreover, Prescott et al. (30) were able to demonstrate that enucleated cells and cell fragments, made from L-cells using cytochalasin B, support vaccinia uncoating and the ' Present address: Institute of Microbiology, University of

Cagliari, Italy.

formation of aggregates which show DNA synthesis. However, Pennington and Follet (29) were unable to detect the formation of mature vaccinia virions in enucleated cells. There are also experiments which suggest a nuclear involvement in vaccinia biosynthesis; Reich and Franklin (31) demonstrated that mitomycin C-treated L-cells, whose nuclei synthesize no DNA at the time of vaccinia infection, are unable to support vaccinia DNA replication and infectious virus formation. Under the same conditions Mengo virus replicated its RNA and produces infectious progeny. Cells that are blocked in metaphase by vinblastin are also unable to support vaccinia DNA synthesis and virus replication (24). Under the same conditions, frog virus DNA and polio virus continue to replicate normally in the cytoplasm (14, 23). Walen (36), by autoradiography, found that the DNA of the infecting vaccinia particles is first associated with the cell chromosomes and that, at later times, intranuclear foci of thymidine incorporation appear to move from the nucleus to the cytoplasm and become part of typical cytoplasmic virosomes. The host DNA is degraded during the first hours after vaccinia infection, and part of this degraded DNA eventually appears in cytoplasmic foci associated with the viral factories (26). Citarella and co-workers (5) found that relatively large amounts of a new vaccinia-induced DNA po-

305

306

J. VIROL.

LACOLLA AND WEISSBACH

layered onto 25 ml of a 1 M sucrose solution made in buffer 3 and centrifuged at 1,000 x g for 7 min. The examination of the nuclear pellet after this last purification step with the phase-contrast microscope and by electron microscopy of thin sections showed no intact cells or visible contamination from cytoplastic debris, intact virus, or vaccinia-specific aggregates. Dahl and Kates (6, 7) have previously reported that much higher centrifugal forces are required to sediment the cytoplasmic vaccinia aggregates. DNA extraction from purified virions. Purified virions were suspended at 20 absorbancy units at 260 nm per ml in 50 mM Tris-hydrochloride (pH 7.8) containing 0.15 M NaCl and 10 mM EDTA. After addition of sodium deoxycholate (to 0.5%) and Pronase (to 1 mg/ml), the mixture was incubated 6 h at 37 C. The solution was extracted three times with MATERIALS AND METHODS phenol-chloroform (50:50) at 4 C, and the DNA was Cells and virus. HeLa S3 cells were grown in precipitated with 2 volumes of ethanol at 20 C. The suspension cultures in F-13 medium (Grand Island DNA was redissolved in 10 mM NaCl, 10 mM EDTA, Biological Co.) supplemented with 5% fetal calf serum and 5 mM Tris (pH 7.4) and treated with pancreatic (Gibco). The WR strain of vaccinia virus, grown in RNase (20 ug/ml) for 30 min at 37 C. The solution was HeLa S3 cells, was used. The viral stock preparations, extracted several times with phenol-chloroform and purified according to Joklik (15), had a ratio of PFU to then dialyzed against 0.015 M NaCl-0.0015 M sodium virus particles of 1 to 100 as determined by optical citrate. DNA extraction from purified nuclei or density and plaque assay. Mode of infection. HeLa S3 cells, growing logarith- cytoplasm. Purified nuclei were suspended at 107 mic phase, were collected by centrifugation and nuclei/ml in 50 mM Tris-hydrochloride (pH 7.8) suspended at a concentration of 107 cells/ml in F-13 containing 0.15 M NaCl and 10 mM EDTA and medium supplemented with 5% heat-inactiP Ated fetal incubated 6 h at 37 C with 0.5% sodium dodecyl calf serum and 1% lactalbumin hydrolyzate (Gibco). sulfate and Pronase (1 mg/ml). DNA was then isoVaccinia virus was added to the culture (10 PFU/ lated from the solution as described for the vaccinia cell) and allowed to adsorb in the presence of 20 mM virions. Cytoplasmic fractions were brought to the salt MgCl2 for 50 min at 37 C. The cells were then washed concentrations given above for the nuclear DNA twice to remove the unadsorbed virus, resuspended at extraction, and the same procedure was then fol106 cells/ml in the above growth medium, and incu- lowed. DNA synthesis in purified nuclei. Nuclei were bated at 37 C. The rate of DNA synthesis was followed by addition prepared by the detergent washing procedure of of 1 mCi of [methyl-3H]thymidine per ml (20 Ci/ Berkowitz (1) and incubated in a mixture (1 ml) conmmol) to cell cultures and measuring incorporation taining Tris-hydrochloride buffer (pH 8.3), 50 mM; into acid-insoluble counts. MgCl2, 4 mM ATP, 5 mM; dithiothreitol, 2.5 mM; Cell fractionation. Samples of 108 infected, or KCl, 20 mM; dATP, dCTP, dGTP, 0.04 mM; [3H]uninfected, HeLa cells, harvested at various times, TTP, [3H]dGTP, and [H3]dATP with a final specific were washed twice in 0.32 M sucrose, 2 mM MgCl2, activity of 900 counts/min/pmol and 5 x 107 nuclei. The incubation was performed at 37 C for 20 min, and 1 mM KPO4 (pH 7.3) (buffer 1) and collected by centrifugation at 500 x g for 3 min. The washed cells followed by chilling to 0 C and isolation of the nuclei then were processed by the following modification of from the assay mixture by centrifugation. DNA was the Berkowitz procedure (1) for the isolation and the prepared from the nuclei as described above. DNA-DNA hybridization. Membrane filters conpurification of the nuclei. The cells, at a concentration of 2 x 107 cells/ml, taining vaccinia DNA (5 ug/filter), extracted from were allowed to swell at 0 C for 5 to 10 min in 10 mM purified virus, or HeLa DNA (50 Ag/filter), extracted NaCl, 5 mM EDTA, and 1 mM KPO4 (pH 7.3) (buffer from nuclei of uninfected cells, were made according 2) and then broken in a Dounce homogenizer to give to Gillespie and Spiegelman (10) using Sartorius more than 95% cell disruption. Centrifugation at 500 nitrocellulose membranes SM11306. The DNA-DNA x g for 3 min was used to remove the crude nuclei annealing method was as described by Denhardt (9). from the cytoplasmic fraction. The crude nuclear The input 3H-labeled DNA was denatured by incupellet was further purified by suspension of 108 nuclei bating in 0.3 N NaOH at 37 C for 15 min. After in 5 ml of 0.32 M sucrose plus 1 mM KPO4 (pH 7.3) cooling in ice, the alkaline solution was neutralized by containing 0.3% Triton N101 (Ruger Chemical Co.) addition of Tris-hydrochloride (pH 7.4) to a final (buffer 3) followed by Dounce homogenization. The concentration of 5 mM and 1 N HCI as required. nuclei were pelleted by centrifugation at 500 x g for Hybridizations were carried out at 65 C for 36 h in 3 min, and suspended in buffer 3. After Dounce glass scintillation vials containing 1 filter/vial in 1 ml homogenization to break any clumps, the nuclei were of 0.02% (wt/vol) each of bovine serum albumin,

lymerase were found associated with the cell nuclei, although it remained undetermined as to what part of this association was an artifact caused by adsorption of vaccinia DNA polymerase from the cytoplasm. We have reinvestigated the possible roll of the cell nucleus as a site for vaccinia DNA synthesis and present evidence that part of vaccinia DNA replication normally occurs in the nucleus. Furthermore, this nuclear involvement may be necessary for the synthesis of mature virions. The synthesis of vaccinia DNA can also be demonstrated in vitro with purified, intact nuclei isolated from infected cells.

-

VOL. 15, 1975

VACCINIA VIRUS INFECTION OF HeLa CELLS.

307

I.

ficoll, and polyvinyl pyrrolidone in 3 x SSC. SSC is a solution of 0.15 M sodium chloride plus 0.015 M sodium citrate. At the end of the incubation the filters were washed by suction filtration with 50 ml/side of 2 x SSC, dried, and counted for the determination of the bound 'H-labeled DNA.

DNA and vaccinia DNA filters. Using the conditions described above, from 48 to 55% of the vaccinia DNA annealed to vaccinia DNA filters, and up to 20% of the HeLa DNA attached to HeLa DNA filters in 36 h. Table 1 also shows that the labeled DNA obtained from the cytoplasm of 3-h vaccinia virus-infected cells, incubated with [3H ]thymiRESULTS dine after infection, is almost exclusively vacThe interpretation of the experiments in this cinia DNA as measured by DNA-DNA hybridization. This would be expected on the basis of paper depend on (i) the ability to distinguish vaccinia DNA from HeLa cell DNA by DNA- earlier reports (4, 6, 7, 16, 20). A small amount DNA molecular hybridization techniques and of hybridization to HeLa DNA filters is ob(ii) the preparation of nuclei which are devoid of served with the cytoplasmic DNA from infected cytoplasmic contaminants. Both of these points cells. This may indicate that some host DNA has been released from broken nuclei into the are discussed below. DNA-DNA hybridization. The use of DNA- cytoplasm or that mitochondrial DNA synthesis DNA hybridization to detect vaccinia or HeLa is occurring. It was surprising, however, to find DNA is illustrated in Table 1. The results that the nuclei of these vaccinia-infected cells clearly show that the hybridization of vaccinia seemed to contain appreciable quantities of DNA to itself is unaffected by the presence of labeled vaccinia DNA in addition to the labeled large amounts of HeLa DNA. Similarly, the host DNA. As stated above, these nuclei were extensively purified by washing and sedimentapresence of vaccinia DNA has no measurable effect on the self-hybridization of HeLa DNA in tion in detergent-containing solutions and were free of visible cytoplasmic or viral contaminaour studies. The cross-hybridization of vaccinia DNA to HeLa DNA is less than 0.2% when tion in the electron or light microscope. Nonadsorption of cytoplasmic vaccinia measured with either 3H-labeled vaccinia DNA and HeLa DNA filters or with 'H-labeled HeLa DNA by nuclei. Since part of the pulse-labeled TABLE 1. DNA-DNA hybridization studies Source of DNA

'H-labeled DNA from purified vaccinia virus 'H-labeled DNA from purified vaccinia virus 3H-labeled DNA from purified vaccinia virus

Input

(Mg)

3HHeLa DNA Vaccinia DNA filter (5 ug) filter (50 ug) labeled DNA (counts/ Counts/ % Hy- Counts/ % Hymin bridized min bridized min)

2,235 4,910 9,660 19,680

55.6 48.9 48.1 49.0

20,080

9,600

20,080

9,695

47.8 48.3

9,640 9,750

48.0 48.5

3H-labeled DNA from purified vaccinia virus

0.2 0.5 1.0 2.0

4,015 10,040 20,080 40,160

'H-labeled DNA plus unlabeled HeLa DNA (1Mgg) 'H-labeled DNA plus unlabeled HeLa DNA (10 Mg) 'H-labeled DNA plus unlabeled HeLa DNA (50 Mg) 'H-labeled DNA plus unlabeled HeLa DNA (5100 g)

1.0 1.0 1.0 1.0

20,080 20,080

30

1.0 .5 5.0 10.0

46,370 115,925 231,850 463,700

9,415 20,285 37,095 69,090

20.3 17.5 16.0 14.9

9,230 9,600 9,505 9,410

20.7

'H-labeled DNA from uninfected HeLa cells 'H-labeled DNA from uninfected HeLa cells 'H-labeled DNA from uninfected HeLa cells 'H-labeled DNA from uninfected HeLa cells 'H-labeled DNA plus unlabeled vaccinia DNA (0.1 Mg) 'H-labeled DNA plus unlabeled vaccinia DNA (1.0 Mg) 'H-labeled DNA plus unlabeled vaccinia DNA (10.0 ug) 3H-labeled DNA plus unlabeled vaccinia DNA (50.0 Mg)

1.0 1.0

46,370 46,370

1.0 1.0

46,370

'H-labeled DNA from vaccinia infected cells' Cytoplasmic DNA Nuclear DNA Nuclear DNA Nuclear DNA

0.8 1.4 2.8 5.5

9,800 5,125 10,250 20,500

46,370

65

200 720

1,255 2,180

0.17

0.15

560

.12

20.5 20.3

75

.16

2.0 14.0 12.2 10.6

4,860 1,115 2,180 4,305

19.9

aVaccinia-infected cells were exposed to 1 MCi of [3H Ithymidine per ml from 150 to 180 min postinfection.

49.6 21.8 21.3 21.0

308

LACOLLA AND WEISSBACH

vaccinia DNA is found in the cytoplasm, it was necessary to exclude nonspecific adsorption as the source of the vaccinia DNA found in the purified nuclei. Another trivial possibility to be excluded was that the vaccinia cytoplasmic aggregates (virosomes) would cosediment with the nuclei in the purification procedure. To test these possibilities, we mixed 108 crude unlabeled nuclei, obtained from either normal cells or cells infected for 3 h with vaccinia virus, with [3H ]thymidine-labeled cytoplasm obtained from 3 h vaccinia-infected cells which contained 1.5 x 106 counts/min. The nuclei after reisolation and purification from these mixtures by our standard technique, contained less than 0.1% of the 3H counts originally present in the infected cytoplasm. Brief sonic treatment of the [3H ]thymidine-labeled cytoplasm from the infected cells to break virosomes and release the viral DNA before incubation with the nuclei did not alter the results. Nuclei were also incubated, in the same way, with 3H-labeled DNA extracted from purified vaccinia virus (1 x 106 counts/min total) or 3H-labeled DNA purified from uninfected HeLa cells (2 x 106 counts/min total). The nuclei

J. VIROL.

isotope found in nuclei of labeled infected cells by two- to fourfold. In fact, no dilution was observed. Thus, our control experiments failed

to demonstrate any artifactual contamination of the purified nuclei with vaccinia virosomes or vaccinia DNA from the cytoplasm of infected cells. However, one cannot exclude the presence of specific viral DNA replication complexes, not revealed by electron microscopy, which are located on the exterior of the nuclear membrane. Synthesis of vaccinia DNA in infected cells. To investigate the participation of the host cell nucleus in vaccinia DNA replication, HeLa cells were infected with vaccinia virus (10 PFU/cell) and DNA synthesis was measured hourly in both the cytoplasm and nucleus. The rate of DNA synthesis, as measured by [3H]thymidine incorporation in a 10-min pulse, showed an initial depression and then increased and reached a maximum to 2 to 3 h after the initial 1-h adsorption period of virus. This DNA synthesis occurred both in the cytoplasm and nucleus (Table 2, Fig. 1A). Uninfected (mockinfected) cells show 95% of the [3H]thymidine label in the nucleus. Fig. 1A shows that the rate isolated from these mixtures again contained of DNA synthesis in mock-infected cells reinsignificant amounts of radioactivity. In addi- mains constant throughout the experimental tion, other pgrification procedures used to ob- period. tain nuclei were tested (3, 28, 32), and the In the cytoplasm of infected cells, the rate of results obtained were the same as those found [3H]thymidine incorporation into DNA reaches with our standard procedure both in the total its maximal level about 3 h postinfection. This amount of 3H-labeled DNA (cytoplasmic or DNA is almost exclusively viral DNA, as shown nuclear) recovered and in the hybridization by the hybridization data in Table 1. At its results. maximal level the rate of viral DNA synthesis is We have also carried out dilution experiments three times higher than the rate of total DNA with vaccinia-infected HeLa cells which were synthesis in uninfected cells. By 4 h postinfecpulse labeled for 20 minutes with [3H ]thymi- tion, the rate of viral DNA synthesis has dedine beginning at 3 h after infection. These creased sharply (Table 1, Fig. 1A). labeled cells were divided into portions and The data in Table 2 and Fig. 1A also show mixed with various amounts of cytoplasm ob- that the rate of DNA synthesis in the nuclei of tained from nonlabeled 3-h infected cells. The infected HeLa cells roughly parallels that obnuclei were then purified from [3H ]labeled in- served in the cytoplasm. There is an initial fected cells, in the presence or absence of exog- decrease of [3H]thymidine incorporation at 1 h enous vaccinia-infected cytoplasm, and their postinfection to 40% the rate shown by control DNA was examined by DNA-DNA hybridiza- cells. The DNA synthesis then increases tion. The nuclei obtained from labeled infected so that by 3 h postinfection it is higher cells which had been broken in the absence of sharply, than the nuclear DNA synthesis of uninfected exogenous cytoplasm or in the presence of a cells. Within 1 h thereafter the rate of DNA two- or fourfold excess of unlabeled cytoplasm synthesis in the nuclei of infected cells has showed, in each case, the same total amount of decreased to less than 30% the rate of unin3H-labeled DNA (2.7 i 0.2 x 106 counts/min) fected cells. Since the inhibition of host DNA and the same percentage of vaccinia specific synthesis by vaccinia infection has been reDNA (50%). If the nuclear content of vaccinia ported by Jungwirth and Launer (17) and Joklik DNA was a trivial contaminant from the cyto- and Becker (16), the presence of relatively large plasm, the addition of cytoplasm from unlabeled amounts of pulse-labeled DNA in the nucleus of vaccinia-infected cells should have diluted the vaccinia-infected cells was, as mentioned be-

309

VACCINIA VIRUS INFECTION OF HeLa CELLS. I.

VOL. 15, 1975

TABLE 2. DNA Synthesis in Vaccinia-Infected HeLa Cellsa Hybridization of nuclear DNA Culture

Fraction

Uninfected cells

Total counts/min x 10-6

%

0.7 13.8

4.8 95.1

14,134

15.3

0

Cytoplasm Nuclei

Infected cells 1h 2h 3h 4h

5h

HeLa DNA filter Input DNA (counts/min) (% counts/min hybridized)

Vaccinia DNA filter (% counts/min hybridized)

Cytoplasm Nuclei

1.04 4.16

19.9 80.0

2,480

16.3

0.4

Cytoplasm Nuclei

33.6 8.42

79.8 20.1

4,240

9.9

20.6

Cytoplasm Nuclei

28.8 13.2

68.5 31.4

5,740

5.8

36.8

Cytoplasm Nuclei

11.4 3.62

75.7 24.2

2,730

6.3

31.5

Cytoplasm Nuclei

5.41 2.8

65.8 34.1

1,836

9.8

26.6

a HeLa cells were incubated with 10 PFU of vaccinia virus per cell. After 1 h of adsorption at 37 C, unadsorbed virus were washed out, and the cells, suspended in F-13 medium at 5 x 105 cell/ml, were reincubated at 37 C (zero time). At 1, 2, 3, 4, 5 and 6 h after the end of virus adsorption, 108 cells were labeled for 10 min with 1 gCi of [methyl-'HJthymidine per ml (specific activity, 2 Ci/mmol) and then immediately chilled by the addition of an equal volume of cold F-13 medium followed by centrifugation at 0 C. The uninfected cell result was obtained with cells which were mock infected under the same conditions. Fractionation of cells, nuclear DNA extraction, and hybridization conditions were as described. Input DNA for hybridization was 10 ug per sample. The percentage of counts per minute values for the hybridization data have been corrected for blank filters which contained no DNA. These blanks were run in each experiment and averaged 35 counts/min.

z

40 3.0

X

B

-z

0~~~~~~~~~ 080 2

2.0 1.0O

be 2o 3 4 40- I23

-

0

4

~~~~~~~~20

x

1 23 45 HOURS

0

12 34 5 HOURS

FIG. 1. (A) Rate of [3H]thymidine incorporation into cytoplasmic and nuclear fraction of uninfected and infected HeLa cells. Symbols: 0, nuclei from uninfected cells; A, cytoplasm from uninfected cells; 0, cytoplasm from infected cells; 0, purified nuclei from infected cells. (B) Graphic representation of the percentage of the hybridization of the DNA extracted from the purified nuclei of infected cells. Symbols: 0, with HeLa DNA; A, with vaccinia DNA. Experimental conditions were as described in footnote to Table 2. The DNA-DNA hybridization conditions were as described. The data indicate the fraction of all

fore, unexpected. To ascertain the nature of the pulse-labeled DNA in these nuclei, the labeled DNA was hybridized against vaccinia or HeLa DNA-containing filters. Table 2 shows that both HeLa and vaccinia DNA are synthesized in the nuclei of infected cells. By 3 h postinfection, approximately 65% of the pulse-labeled nuclear DNA is vaccinia DNA. This is also shown in Fig. 1B, in which the hybridization data shown in Table 2 have been normalized to the hybridization efficiency obtained with HeLa DNA filters and known 3H-labeled HeLa DNA (15.3%, see Table 2) or with vaccinia DNA filters and 3H-labeled vaccinia DNA from purified virions (55.6%, see Table 1). The hybridization data of Table 2 are expressed as a perradioactivity in the infected pulse-labeled nuclei which represents either HeLa or vaccinia DNA. The results have been normalized to control HeLa DNA-HeLa DNA and vaccinia DNA-vaccinia DNA hybridizations and are expressed as percentages of all control values.

J. VIROL.

LACOLLA AND WEISSBACH

310

centage of these maximal hybridization results and plotted in Fig. 1B. The figure illustrates the fraction of labeled DNA isolated from purified infected nuclei which represents HeLa DNA and vaccinia DNA. The reciprocal relationship between host and viral DNA that one would anticipate from this type of plot is observed. As the rate of vaccinia DNA synthesis increases in the period 2 to 4 hours after infection the rate of synthesis of host DNA in the nucleus initially drops. From Table 2 it can be calculated that the host DNA synthesis drops to 30% the rate of uninfected control cells by 1 h postinfection and remains at that level for the next 2 h. At later times both the viral and host DNA synthesis drop markedly. Vaccinia DNA synthesis in FUdR-treated cells. Treatment of cells with 5-fluorodeoxyuridine (FUdR) leads to a cessation of DNA synthesis which can be readily reversed by thymidine (33). Cells infected with vaccinia virus in the presence of FUdR will synthesize neither host nor vaccinia DNA, but will express and accumulate early viral functions not dependent upon viral DNA synthesis (19). The effect of accumulating early viral products in FUdR-blocked vaccinia-infected cells on the subsequent synthesis of DNA when the FUdR block is removed is shown in Table 3 and Fig. 2. In these experiments, infected (or control cells)

incubated in the presence of FUdR and pulse labeled with 5 x 10-I M [3H ]thymidine at various times after vaccinia infection. This amount of thymidine is sufficient to overcome the FUdR block and permit initiation of DNA replication in the cells. The rate of DNA synwere

0

O

0

z u0

z0 Z

, 0

z

-

Z

sD

v

I-1

0

2

4 3 HOURS

5

0

2

4 3 HOURS

5

FIG. 2. DNA synthesis in FUdR-treated cells. The conditions for this experiment are described in the footnote to Table 3. Mock infected cells received the same manipulations but without viral infection. (A) Symbols: A, counts per minute incorporated into cytoplasm of mock infected cells; 0, into the nucleus of mock-infected cells; *, into cytoplasm of infected cells; 0, into nuclei of infected cells. (B) Percentage of hybridization of "H-labeled DNA extracted from the purified nuclei obtained from the vaccinia-infected cells (see Table 3). The data is expressed as described in the legend to Fig. 1B. TABLE 3. Vaccinia DNA synthesis in FUdR-treated cellsa Hybridization of nuclear DNA

Culture

FUdR-treated and infected cells 1h 2h 3h 4h

5h

HeLa DNA filter (counts/min) (% counts/min hybridized)

Vaccinia DNA filter (% counts/min hybridized)

Fraction

Total counts/min

%

Cytoplasm Nuclei

0.13 x 106 2.48 x 106

5.0 94.9

11,345

13.2

1.7

Cytoplasm Nuclei

1.16 x 106 2.58 x 106

30.9 69.0

11,870

7.7

9.3

Cytoplasm Nuclei

2.63 x 106 3.74 x 106

41.2 58.7

15,393

6.9

22.3

Cytoplasm Nuclei

5.28 x 106 8.15 x 106

39.3 60.6

34,297

7.0

22.7

Cytoplasm

3.21 x 106 4.30 x 106

42.7 57.2

18,176

9.6

19.8

Nuclei

Input DNA

a HeLa S3 cells were infected with vaccinia virus at 10 PFU/cell in the presence of 10-I M FUdR. After the 1-h adsorption period, the cells were washed and incubated in F-13 medium containing 10-s M FUdR and 10-5 M uridine. At 1, 2, 3, 4 and 5 h postinfection, 10" cells were labeled for 10 min with [9H]thymidine (1 ,Ci/ml, 5 x 10-I M) and immediately chilled. Fractionation of cells, nuclear DNA extraction, and hybridization conditions were as described. Input DNA for hybridization was 5 ,g per sample.

VOL. 15, 1975

311

VACCINIA VIRUS INFECTION OF HeLa CELLS. I.

thesis in control, uninfected cells rises slightly over a period of 5 h (Fig. 2A). Infected cells show a marked burst of DNA synthesis which reaches a maximum at 4 h postinfection and then drops sharply. Under these conditions, the bulk of the newly synthesized DNA in the vaccinia-infected cell is found in the nucleus. This may be contrasted to the results obtained in a normal infection where the nuclei contain only 30% of the total labeled DNA found in the cell (Table 2, Fig. 1A). DNA hybridization studies (Table 2, Fig. 2B) again show that a significant proportion (about 40%) of the pulse-labeled nuclear DNA is vaccinia specific at 3 to 4 h after infection. As with the normal infection, the rate of host DNA synthesis initially drops in the infected cell and then remains at a reduced level during the synthesis of the vaccinia DNA. Approximately 80% of the newly synthesized DNA is accounted for in the studies of Fig. 2B. Synthesis of viral DNA in 2-min pulse experiments. The experiments listed above utilized 10-min pulses with [3H ]thymidine, and there was the possibility that part of the newly synthesized viral DNA in the nucleus represented a transfer from DNA initially made in the cytoplasm. Further tests with shorter pulse times indicate this is not the case. A 2-min pulse of [3H]thymidine at 4 h after infection in the presence of FUdR results in 58% of the newly synthesized DNA being found in the nucleus (Table 4). A large part of this DNA is vaccinia DNA, and the data resembles that shown in Table 3 for the 10-min pulse. This data would tend to eliminate the rapid transfer of cytoplasmically synthesized vaccinia DNA into the nucleus as an explanation for the data reported herein, but does not completely exclude this possibility. Conversely, it would also indicate that the vaccinia DNA found in the

cytoplasm is not derived from the nucleus and that the synthesis of viral DNA occurs in both compartments.

Synthesis of vaccinia DNA in isolated nuclei. Purified, detergent-washed nuclei, isolated from vaccinia-infected cells, also are capable of synthesizing vaccinia DNA in vitro. Such nuclei, incubated with 'H-labeled deoxynucleoside triphosphates under the proper conditions as described above produce newly synthesized DNA which is predominantly vaccinia DNA. Approximately 95% of the 'H-labeled DNA made by the isolated nuclei contains vaccinia DNA sequences which hybridize to a vaccinia DNA filter (Table 5). The isolated nuclei from vaccinia-infected cells thus reflect the in vivo findings in which both vaccinia and HeLa DNA synthesis is observed to occur in the nucleus. The isolated nuclei, in fact, show a higher percentage of vaccinia DNA synthesis than found in vivo and may represent a useful system for further studies in this area. Control experiments, in which the same number of nuclei (5 x 107) obtained from uninfected cells, were incubated with 5 gg of exogenous vaccinia DNA under the same conditions showed little de novo vaccinia DNA synthesis (5% of the total). Vaccinia infection in mitomycin C-treated cells. To further investigate the possible role of the cell nucleus in vaccinia DNA synthesis, we followed viral DNA synthesis in cells in which the host nuclear DNA synthesis is shut off by treatment with mitomycin C (34, 35). The synthesis of the viral DNA of frog virus 3, which replicates in the cytoplasm, has been reported not to be inhibited by this drug (23). The effect of mitomycin C on vaccinia virus replication has been investigated by Kit et al. (21), who found that cells, infected for 1 to 2 h (in the presence of 15 gg of mitomycin per ml), synthesized DNA at about 20% the rate of mitomycin-

TABLE 4. Two-minute [3H]thymidine pulse of vaccinia-infected HeLa cellsa Hybridization of nuclear DNA Fraction

Total counts/min

%

Cytoplasm

6.85 x 10'

41.9

Nuclei

9.48 x 10'

58.0

DNA

HeLa DNA filter (% counts/min

(counts/min)

hybridized)

Vaccinia DNA filter (% counts/min hybridized)

1,540

11.3

34.9

Input

a HeLa S3 cells were vaccinia infected (10 PFU/cell) in the presence of FUdR and incubated at 37 C in F-13 medium containing 10-' M FUdR and 10-5 M uridine. At 4 h postinfection, 2 x 106 cells were labeled for 2 min with [methyl-HjLhymidine (1 IACi/ml, 5 x 10-7 M), followed by the addition of 5 x 10-4 M unlabeled thymidine, and the cells were immediately collected by centrifugation at 0 C. Input DNA for hybridization was 10 ;ig per sample.

312

LACOLLA AND WEISSBACH

treated, uninfected cells. Reich and Franklin (31) found that mitomycin inhibited the production of infectious virus and DNA synthesis at 6 to 8 h after infection. We have examined vaccinia DNA and virus formation in cells which have been exposed to mitomycin C followed by -moval of the antibiotic prior to infection with 1irus. freatment of growing HeLa cells with 10 jig of mitomycin C per ml causes a rapid decrease in DNA synthesis with a minor inhibition of either

J.- VIROL.

RNA or protein synthesis. After 5 h of treatment with mitomycin C, DNA synthesis is inhibited by 95%, whereas RNA and protein synthesis are inhibited 10 and 20%, respectively. If, at this point, the cells are removed from the mitomycin, placed in fresh medium, and then infected with vaccinia virus, synthesis of vaccinia DNA, as measured by DNA-DNA hybridization, is again found to occur both in the nucleus and cytoplasm (Table 6). However, the rate of vaccinia DNA synthesis in the nucleus or cyto-

TABLE 5. In vitro synthesis of host and vaccinia DNA by isolated nucleia Input DNA

'H-labeled DNA source Vaccinia virus HeLa cells In Vitro nuclei from vacciniainfected cells

lAg

Counts/min

1.1 0.4 10

63,922 5,482 15,375

Hybridization (%) Vaccinia DNA filter HeLa DNA filter (100 JAg) (5 Ag)

35 0.04 33.7

0.07 25 4.7

aThe incubation of purified nuclei with 3H-labeled deoxynucleoside triphosphates was as described. The nuclei were obtained from HeLa cells 6 h after infection with vaccinia virus at a multiplicity of 10 PFU/cell. The total amount of 3H-labeled DNA synthesized in this system was 4,700 pmol of nucleotide incorporated per 5 x 107 nuclei per 20 min. After the incubation the DNA was extracted from the nuclei as described and hybridized to membrane filters containing either vaccinia or HeLa DNA. TABLE 6. Vaccinia infection of cells pretreated with mitomycin Ca Hybridization of nuclear DNA Culture

Uninfected cells Oh

Infected cells 1h 2h 3h 4h 5h

Fract ion

HeLa Vaccinia Input DNA DNA filter DNA filter (counts/min) (% counts/min (% counts/min hybridized) hybridized)

Total counts/min

Cytoplasm Nuclei

3,82 x 103 5.97 x 10'

6.0 93.9

4,861

4.4

Cytoplasm Nuclei

1.96 x 10' 2.57 x 10'

43.2 56.7

2,773

4.8

9.5

Cytoplasm Nuclei

1.22 x 105 4.83 x 10'

71.6 28.3

3,671

3.2

25.4

Cytoplasm Nuclei

1.28 x 10' 5.33 x 10'

70.6 29.3

4,695

2.7

25.4

Cytoplasm Nuclei

1.87 x 10' 1.04 x 10'

64.2 35.7

3,890

2.9

22.0

Cytoplasm Nuclei

7.48 x 10' 3.37 x 10'

68.9 31.0

2,878

2.9

28.7

HeLa S3 cells, pretreated for 5 h with 10 ug of mitomycin C, were infected with vaccinia virus at 10 PFU/cell for 1 h at 37 C. The unadsorbed virus was removed by washing, and the cells were reincubated at 37 C in F-13 medium. At 1, 2, 3, 4 and 5 h postinfection, 108 cells were labeled for 10 min with [methyl-'H]thymidine at 1 ,uCi/ml and then chilled. The 0 h control consisted of 10' mitomycin C-pretreated but uninfected cells. Fractionation of cells, nuclear DNA extraction, and hybridization conditions were as described. Input DNA for hybridization was 150 ,g per sample, and the percentage of hybridization is corrected for the blank values. a

VOL. 15, 1975

VACCINIA VIRUS INFECTION OF HeLa CELLS. I.

plasm in these mitomycin-pretreated cells (Table 6) is about 5 to 10% of that found in untreated, infected cells (Table 2). The rate of host DNA synthesis is decreased to the same extent. Because of the low levels of [3H ]thymidine incorporated by mitomycin-treated cells, it was necessary to use large amounts of DNA (150 jg) in the hybridization experiments. Under these conditions the efficiency of HeLa DNADNA hybridization is decreased to 4 to 5% (Table 6). Nevertheless, it is clear that even under conditions in which the synthesis of both host and viral DNA are markedly inhibited, a significant proportion of the newly synthesized vaccinia DNA is located in the nucleus. Though total viral DNA synthesis in these pretreated cells is inhibited 95%, the kinetics of vaccinia DNA synthesis in the cytoplasm and nucleus parallels the synthesis in the normal infection. We have also assayed the number of infectious virus produced after 24 h by cells pretreated with mitomycin C for 5 h before infection, and found it to be 0.4% of that produced by untreated control cells. Thus, pretreatment of host cells to stop host nuclear DNA synthesis has a profound effect on subsequent replication of vaccinia DNA and production of mature virus. DISCUSSION The results in this study suggest that both the nucleus and cytoplasm are involved in the replication of vaccinia virus and that host DNA synthesis is only partially suppressed during viral development. The latter point is in agreement with the data of Jungwirth and Launer (17), who found that cell DNA synthesis in vaccinia-infected cells is reduced about 75% 6 h after infection. This residual host cell DNA synthesis was unaffected by the multiplicity of infection over a range from 40 to 400 virus particles per cell. Synthesis of nuclear DNA in vaccinia-infected cells was also observed by Cairns (4) in his autoradiographic studies, which visualized centers of virus synthesis in the cytoplasm. Further autoradiographic studies by Walen (36) have suggested that initial DNA replication in vaccinia virus-infected cells is associated with the chromosome. These studies could not distinguish between host and viral DNA synthesis. The use of DNA-DNA hybridization in this paper permits the unequivocal distinction between these two types of DNA. There have been other reports that a functioning nucleus is necessary for vaccinia replication. Miller and Enders (22) showed that leukocytes could not replicate vaccinia virus unless the leukocytes were stimulated by

313

phytohemagglutinin. The same phenomenon has been observed with herpes simplex virus, which is known to replicate its DNA in the nucleus (25). This is consistent with the report of Marcus and Robbins (24) that metaphase cells cannot support vaccinia replication. The extensive occurrence of vaccinia DNA in the nucleus of infected cells does not seem to be an artifact occurring during breakage of the cells. Thus, the mixing and reconstruction experiments using unlabeled nuclei from normal or vaccinia-infected cells and [3H ]thymidinelabeled cytoplasm from infected cells do not show adsorption of vaccinia DNA from the cytoplasm by the nucleus. Furthermore, the percentage of pulse-labeled vaccinia DNA found in the nuclei of vaccinia-infected cells can range from 10 to 100% of that found in the cytoplasm (Tables 2 and 3), depending upon the time and conditions of pulse labeling. Even the shortest pulse label experiments (2 min). show extensive synthesis of vaccinia DNA in the nucleus. Since it is difficult to imagine the movement of large amounts of vaccinia DNA from the cytoplasm to the nucleus in this short time span, we take this as evidence that vaccinia DNA synthesis occurs both within the nucleus and cytoplasm. Nuclear synthesis of vaccinia DNA in infected cells is reflected in the in vitro studies with isolated nuclei obtained from these cells. Such isolated nuclei show significant de novo synthesis of vaccinia DNA and, in fact, produce more viral than host DNA sequences. That this, in turn, is relevant to in vivo events is indicated by the observation that nuclei isolated from uninfected cells, when mixed with exogenous vaccinia DNA, fail to synthesize significant amounts of new vaccinia DNA. These observations, coupled with the finding that cells pretreated with mitomycin C to stop host nuclear DNA synthesis also show a concomitant decrease in subsequent vaccinia DNA synthesis and almost complete abolition of infectious virus formation, might indicate that the nucleus plays some essential role in the replication of vaccinia DNA or mature virus. The possible dependence of vaccinia virus DNA replication on the host nuclear apparatus can be compared to the results of McAuslan and Smith (23) with FV-3, a cytoplasmic virus in which DNA synthesis is unaffected by mitomycin C. Furthermore, these workers showed that absorbed FV-3 DNA can be separated from crude nuclei by simple centrifugation through sucrose. Some vaccinia DNA synthesis, by contrast, does seem to occur in the nucleus, and blockage of the host DNA synthesis will subsequently stop vaccinia DNA synthesis. The relationship

314

LACOLLA AND WEISSBACH

J. VIROL.

of the vaccinia DNA synthesized in the nucleus 10. Gillespie, D., and S. Spiegelman. 1965. A quantitative assay for DNA-RNA hybrids with DNA immobilized on to that made in the cytoplasm remains unclear. a membrane. J. Mol. Biol. 12:829-842. Our data provides no clear evidence of move- 11. Green, M., J. T. Parson, M. Pina, K. Fujinaga, K. ment of vaccinia DNA from one cellular comCaffier, and I. Landgraf-Leus. 1970. Transcription of adenovirus genes in productivity infected and in transpartment to another, although this is an obvious formed cells. Cold Spring Harbor Symp. Quant. Biol. possibility which may be pertinent to the bio35:803-818. synthesis of vaccinia DNA. 12. Groyon, R., and A. J. Kniazeff. 1967. Vaccinia virus The presence of vaccinia DNA in the nucleus infection of sychronized pig kidney cells. J. Virol. 1:1255-1264. may be related to the observation of Kates H. 1962. Factors involved in the initiation of (18) that vaccinia mRNA contains poly(A) 13. Hanafusa, multiplication of vaccinia virus. Cold Spring Harbor sequences. It has been postulated that the Symp. Quant. Biol. 27:209-217. poly(A) sequences at the end of mRNA may be 14. Johnson, T. C., and J. J. Holland. 1965. Ribonucleic acid and protein synthesis in mitotic HeLa cells. J. Cell Biol. involved in transport of mRNA out of the 27:565-574. nucleus. The occurrence of poly(A) sequences in 15. Joklik, W. K. 1962. Preparation and characteristics of vaccinia mRNA presumed to be synthesized highly purified radioactively labeled poxvirus. Bioand utilized solely in the cytoplasm seemed chim. Biophys. Acta 61:290-301. inconsistent with this postulate, even though a 16. Joklik, W. K., and Y. Becker. 1964. The replication and coating of vaccinia DNA. J. Mol. Biol. 10:452-474. poly(A) polymerase has been reported in vac- 17. Jungwirth, C., and J. Launer. 1968. Effect of poxvirus cinia virus (2). However, the synthesis of vacinfection of host cell deoxyribonucleic acid synthesis. J. cinia DNA in the nucleus, as reported here, may Virol. 2:401-408. be accompanied by the synthesis of vaccinia 18. Kates, J. R. 1970. Transcription of the vacccinia virus genome and the occurrence of polyadenylic acid semRNA in the nucleus, which could be expected quences in messenger RNA. Cold Spring Harbor Symp. to contain poly(A) sequences. Quant. Biol. 35:743-752. It is possible that a specific replicative step 19. Kates, J. R., and B. R. McAuslan. 1965. Relationship between protein synthesis and viral deoxyribonucleic occurs in the nucleus and that this step is

for subsequent synthesis and maturation of the virus in the cytoplasm. The nature of these nuclear events remains to be elucidated, but these findings may give clue to the mechanism of vaccinia DNA replication and be related to the process by which the virus inhibits the synthesis of host DNA (13, 17, 27).

necessary

ACKNOWLEDGMENTS We are indebted to Jayme Aucker for her technical assistance. LITERATURE CITED 1. Berkowitz, D. M., T. Kakefuda, and M. D. Sporn. 1969. A simple and rapid method for the isolation of enzymatically active HeLa cells nuclei. J. Cell Biol. 42:851-855. 2. Brown, M., J. W. Dorson, and F. J. Bollum. 1973.

3. 4.

5.

6. 7.

8.

9.

Terminal riboadenylate transferase: a poly A polymerase in purified vaccinia virus. J. Virol. 12:203-208. Busch, H. 1967. Isolation and purification of nuclei. Methods Enzymol. 12:421-448. Cairns, J. 1960. The initiation of vaccinia infection. Virology 11:603-623. Citarella, R. V., R. Muller, A. Schlabach, and A. Weissbach. 1972. Studies on vaccinia virus-directed deoxyribonucleic acid polymerase. J. Virol. 10:721-729. Dahl, R., and J. R. Kates. 1970. Intracellular structure containing vaccinia DNA: isolation and characterization. Virology 42:453-462. Dahl, R., and J. R. Kates. 1970. Synthesis of vaccinia virus "early" and "late." Virology 42:463-472. Darnell, J. E., L. Phillipson, R. Wall, and M. Adesnik. 1971. Polyadenylic acid sequences: role in conversion of nuclear RNA into messenger RNA. Science 174:507-510. Denhardt, D. 1966. A membrane-filter technique for the detection of complementary DNA. Biochem. Biophys. Res. Commun. 23:641-646.

acid syntheses. J. Virol. 1:110-114. 20. Kato, S., M. Takahashi, S. Kameyama, and J. Kamahora. 1959. A study on the morphological and cytoimmunological relationship between the inclusions of variola, cowpox, rabbitpox, vaccinia (variola origin) and vaccinia IHD and a consideration of the term "Guarnieri body." Biken J. 2:353-363. 21. Kit, S., L. J. Piekarski, and D. R. Dubbs. 1963. Effects of 5-fluorouracil actinomycin D and mitomycin C on the induction of thymidine kinase by vaccinia-infected L-cells. J. Mol. Biol. 7:497-510. 22. Miller, G., and J. F. Enders. 1968. Vaccinia virus replication and cytopathic effect in cultures of phytohemagglutinin-treated human peripheral blood leukocytes. J. Virol. 2:787-792. 23. McAuslan, B. R., and W. R. Smith. 1968. Deoxyribonucleic acid synthesis in FV-3-infected mammalian cells. J. Virol. 2:1006-1015. 24. Marcus, P. I., and E. Robbins. 1963. Viral inhibition in the metaphase-arrest cell. Proc. Nat. Acad. Sci. U.S.A. 50:1156-1164. 25. Nahmias, A. J., S. Kibrick, and R. C. Rosan. 1964. Viral leukocyte interrelationships. I. Multiplication of a DNA virus-herpes simplex in human leukocyte cultures. J. Immunol. 93:69-74. 26. Oki, T., Y. Fujiwara, and C. Heidelberger. 1971. Utilization of host cell DNA by vaccinia virus replicating in HeLa cells irradiated intranuclearly with Tritium. J. Gen. Virol. 13:401-413. 27. Parkhurst, J. R., A. R. Peterson, and C. Heidelberger. 1973. Breakdown of HeLa cell DNA mediated by vaccinia virus. Proc. Nat. Acad. Sci. U.S.A. 70:3200-3204. 28. Penman, S. 1966. RNA metabolism in the HeLa cell nucleus. J. Mol. Biol. 17:117-130. 29. Pennington, T. H., and E. A. C. Follet. 1974. Vaccinia virus replication in enucleate BSC-1 cells: particle production and synthesis of viral RNA and proteins. J. Virol. 13:488-493. 30. Prescott, D. M., J. Kates, and J. B. Kirkpatrick. 1971. Replication of vaccinia virus DNA in enucleated L-

VOL. 15, 1975

VACCINIA VIRUS INFECTION OF HeLa CELLS.

cells. J. Mol. Biol. 59:505-508. 31. Reich, E., and R. M. Franklin. 1961. Effect of mitomycin C on the growth of some animal viruses. Proc. Nat. Acad. Sci. U.S.A. 47:1212-1217. 32. Roizman, B., S. Bachenheimer, E. K. Wagner, and T. Savage. 1970. Synthesis and transport of RNA in herpes-infected mammalian cells. Cold Spring Harbor Symp. Quant. Biol. 35:753-771. 33. Shatkin, A. J., and N. P. Salzman. 1963. Deoxyribonucleic acid synthesis in vaccinia virus-infected HeLa

I.

315

cells. Virology 19:551-560. 34. Shiba, S., A. Terawaki, T. Taguchi, and J. Kawamata. 1959. Selective inhibition of formation of deoxyribonucleic acid in Escherichia coli by mitomycin C. Nature (London) 183:1056-1057. 35. Szylbalski, W. 1964. Mitomycin C profiromycin: chemical mechanism of activation and cross-linking of DNA. Science 145:55-58. 36. Walen, K. H. 1971. Nuclear involvement in poxvirus infection. Proc. Nat. Acad. Sci. U.S.A. 68:165-168.