Vol. 17, No. 2 Printed in U.S.A.
JOURNAL OF VIROLOGY, Feb. 1976, p. 402-414 Copyright 0 1976 American Society for Microbiology
DNA Synthesis in Polyoma Virus Infection IV. Mechanism of Formation of Closed-Circular Viral DNA Deficient in Superhelical Turns KENNETH YU AND WILLIAM P. CHEEVERS* Cancer Research Laboratory, University of Western Ontario, London, Ontario, Canada Received for publication 22 July 1975
A marked reduction in the rate of viral DNA synthesis is accompanied by an alteration to the superhelicity of progeny DNA in polyoma virus-infected cells in which protein synthesis has been inhibited by cycloheximide. Viral DNA molecules formed in the presence of cycloheximide consist predominantly of closed-circular monomeric species (referred to as form Ic) characterized by a decreased superhelix density, corresponding to Aao, = 0.0195, as compared to form I DNA by propidium diiodide-cesium chloride isopycnic analysis. Form Ic is synthesized on pre-existing form I templates without the intervention of progeny form I as an intermediate. It is concluded that inhibition of protein synthesis results in the alteration of some process in the closure of daughter DNA that leads to a marked reduction of superhelical turns of progeny molecules. About two-thirds of form Ic molecules return to the form I conformation upon reversal of cycloheximide inhibition by a mechanism independent of DNA replication.
When protein synthesis is inhibited in polyoma virus-infected cells by either cycloheximide (3, 6) or puromycin (2), a viral DNA component is formed that sediments at neutral pH slightly slower than 20S form I viral DNA. Alkaline velocity sedimentation and isopycnic centrifugation in propidium diiodide-cesium chloride show this DNA (form Ic) to consist of monomeric closed-circular molecules with a decreased superhelicity as compared to form I (2, 6). The present study is concerned with the mechanism of formation of Ic viral DNA in cycloheximide-treated cells. It is shown that newly synthesized Ic is made on pre-existing form I DNA templates. Closed-circular progeny DNA of normal superhelix density is not involved as an intermediate in this process. It is concluded that form Ic arises by a cycloheximide-induced alteration of the closure of daughter molecules. Improperly closed progeny DNA may reacquire the normal tertiary structure of form I upon restoration of protein synthesis. This process does not involve DNA replication. (This work was taken from a thesis by K. Y. for submission to the Department of Bacteriology and Immunology, University of Western Ontario, in partial fulfillment of requirements for the Ph.D. degree.) MATERIALS AND METHODS Tissue culture and virus infection. Primary cul402
tures of mouse embryo cells were grown in McCoy 5A medium supplemented with 10% fetal calf serum at 37 C in 5% CO2-95% air. For experiments, primary cells were subcultured into 100-mm Falcon plastic dishes at 15 x 101 cells per dish. Cultures were grown to stationary phase (approximately 40 x 105 cells) and infected with TSP1 polyoma virus (17) at a multiplicity of 30 plaque-forming units per cell as previously described (4). Isotopic labeling and extraction of DNA. Replicating viral DNA was radioactively labeled by exposure of polyoma-infected cultures to medium containing [methyl-3H ]thymidine (TdR) (New England Nuclear; 40 to 52 Ci/mmol) or [2- "C]TdR (Amersham/ Searle; 40 to 60 mCi/mmol). The conditions of labeling are given in individual experiments. For pulsechase experiments, the labeled medium was decanted and the cells were immediately washed twice with prewarmed (37 C) medium containing 2 x 10-i M TdR and then incubated with this medium for the required chase time. For extraction of DNA, cells were rinsed with ice-cold SSC (0.15 M NaCl plus 0.015 M trisodium citrate) and dispersed by 0.1% trypsin in saline citrate solution (16). Cells were then collected by centrifugation, suspended in SSC and lysed by addition of 0.1 volume of 10% sodium dodecyl sulfate in EDTA buffer (0.1 M NaCl-0.001 M EDTA-0.01 M Tris, pH 7.4). Lysates were centrifuged through 30-ml 15 to 30% (wt/wt) sucrose gradients in EDTA buffer containing 0.5% sodium dodecyl sulfate, formed over a 6-ml cushion of 70% (wt/vol) sucrose in EDTA buffer. Centrifugation was at 26,000 rpm at 23 C for 10 h using a Spinco SW27 rotor. Under these conditions approximately 95% of the cellular DNA in the crude lysates is recovered as high-molecular-weight mate-
VOL. 17, 1976
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
403
cycloheximide. Two points are evident: (i) much less viral DNA was synthesized than in untreated cultures (this effect has been shown to result exclusively from limited initiation of new rounds of genome replication [21 ]); and (ii) 20S form I DNA was replaced by a DNA component sedimenting at 16S. This component is form Ic. Form Ic sediments as monomeric closed-circular viral DNA under alkaline conditions (6) and shows a decreased superhelix density by isopycnic analysis in cesium chloride-PDI (Fig. ic) as compared to form I DNA (Fig. lb). The superhelix density of form Ic was determined as follows. Forms I and Ic, labeled with [14C ]TdR and [3H]TdR, respectively, were isolated as described in Fig. la and purified by cesium chloride-PDI centrifugation followed by Sephadex G-50 column chromatography to remove CsCl and PDI. After concentration by ethanol precipitation, purified forms I and Ic were mixed, partially converted to form H by limited digestion with deoxyribonuclease, and centrifuged to equilibrium in cesium chloridePDI. The distance between the bands corresponding to 3H-labeled II and 3H-labeled Ic was compared with that of "4C-labeled II and 14C_ labeled I according to the Eason and Vinograd equation (9) Aa. = 0.1 [(Ar/Ar*) - 1]. Our measurements indicate a shift in superhelix density from I to Ic equivalent to &ao = 0.0195, which corresponds to a decrease in the number of superhelical turns of about two-thirds (13, 18). Origin of form Ic DNA. The purpose of this experiment was to test the possibility that form Ic arises from form I molecules. Replicating form I DNA was labeled for 2 h with [3H]TdR. The fate of these labeled molecules was then determined during a chase for various periods of time with unlabeled TdR in the presence or absence of cycloheximide. Figure 2 shows the cesium chloride-PDI density gradient profiles of viral DNA chased for 4 h. Without cycloheximide (Fig. 2a), [3HIDNA banded with marker "4C-labeled form I as expected. In the absence of protein synthesis, however, a portion of 3H-labeled form I DNA was chased into the position of Ic (Fig. 2b). The quantitative transfer of label from I to Ic during the 4-h chase period in the presence of cycloheximide is shown in Fig. 3b. These data show that about one-third of form I DNA molecules replicated during a 2-h period were diverted to the RESULTS formation of component Ic during a subsequent Characterization of form Ic DNA. Figure la 2-h period in the absence of protein synthesis. This finding was confirmed by the experishows the sedimentation properties at neutral pH of polyoma DNA formed in the presence of ment described in Fig. 4. Infected cells were
rial on the dense sucrose cushion, whereas polyoma DNA components are resolved within the gradient (6-8, 22). Viral DNA was recovered for subsequent analysis as follows. Appropriate neutral sucrose gradient fractions were combined, and the DNA was precipitated with 2.5 volumes of cold 95% ethanol. After overnight storage at -20 C, the precipitate was collected by centrifugation (10,000 x g; 30 min; 0 C) and dissolved in 0.3 M NaCl-0.001 M EDTA-0.01 M Tris, pH 8.1. Dye-buoyant density centrifugation. The superhelix density of closed-circular polyoma DNA was analyzed by equilibrium centrifugation in CsCl gradients containing propidium diiodide (PDI) (11). i4Clabeled form I marker DNA was prepared by labeling infected cultures with ["4C]TdR (0.2 to 0.5 gCi/ml) from 10 to 30 h postinfection. 'H-labeled form Ic marker DNA was prepared from infected cultures labeled with [3H ]TdR (50 IACi/ml) at 28 to 34 h postinfection in the presence of cycloheximide. Test DNA samples were mixed with the appropriate marker and a solution of CsCl in 0.02 M Tris (pH 8.6)-0.02 M EDTA containing PDI to give a final volume of 3.5 ml, an initial density of 1.52 g/cm3, and a final concentration of 500 gg of PDI/ml. These solutions were then centrifuged to equilibrium at 20 C in a Spinco SW40 rotor for 90 h at 22,000 rpm or alternatively in the SW50L or SW50.1 rotor for 48 h at 40,000 rpm. Radioactivity determinations. For measurement of total radioactive DNA, cell cultures were lysed by addition of 0.5% sodium dodecyl sulfate in EDTA buffer. Lysates were then mixed with an equal volume of cold 10% trichloroacetic acid to precipitate the macromolecular fraction. Acid-insoluble material was collected onto Whatman GF/C glass fiber filters and counted in nonaqueous scintillator as previously described (8). Sucrose gradients were collected using an ISCO model 640 fractionator equipped with model UA-2 UV analyzer at 254 nm. Cesium chloride-PDI gradients were collected using a Buchler Instruments fraction collector. Three-drop fractions were collected using constant positive pressure from the top generated by a multispeed transmission (Harvard Apparatus Co., Dover, Mass.) pushing a column of paraffin oil at the rate of 0.408 ml/min. Fractions were mixed with an excess of cold 5% trichloroacetic acid and the acidinsoluble fraction was collected onto membrane (Millipore) or cellulose ester (Gelman) filters (0.45-um pore size). The filters were washed with 5% trichloroacetic acid in 95% ethanol and counted as described above. Alternatively, fractions were mixed with two drops of an aqueous solution of fraction V bovine serum albumin (1 mg/ml) and adjusted to a final concentration of 5% trichloroacetic acid, and the acid-insoluble fraction was collected by filtration onto Whatman GF/C glass fiber papers.
404
J. VIROL.
YU AND CHEEVERS
c 0 a-
on
x
0
tvb
a
x
E
"._
In.
e0
go
c
0 =x 0
-
0-
N.
C
C
0
E
u
0
U
N1-.
lu e
on
0-
0
u
0
10
20
30
40
FRACTION NO.
FRACTION NO.
FIG. 1. Sedimentation and dye-buoyant density centrifugation properties of polyoma DNA synthesized in cycloheximide-treated cells. (a) Velocity sedimentation at neutral pH. Infected cells were pretreated for 90 min at 28 h postinfection with medium with or without cycloheximide (10 jug/ml) and then pulse labeled for I h with ['H]TdR (25 jCi/ml). Cell lysates were prepared by treatment with sodium dodecyl sulfate and centrifuged in neutral sucrose gradients as described. Symbols: 0, Untreated cells; 0, cycloheximide-treated cells. The direction of sedimentation is from left to right in all velocity sedimentation analyses. The indicated fractions in (a) were combined and the DNA was concentrated by ethanol precipitation and analyzed by isopycnic centrifugation in cesium chloride-PDI (rotor SW50.1). (b) Untreated cells. Symbols: 0, ['HJDNA; x, "C-labeled marker form IDNA. (c) Cycloheximide-treated cells. Symbols: 0, [3HIDNA; x, "4C-labeled marker form IDNA. Density increases from right to left in all cesium chloride-PDI gradients.
labeled for 2 h with ["C ]TdR and then chased additional 2 h with medium containing unlabeled TdR and cycloheximide. ['H ]TdR was administered during the last 1.5 h of the chase period to label newly formed component Ic. Figure 4a shows the distribution in cesium chloride-PDI of closed-circular "C-labeled viral DNA before the chase period, relative to that of marker 'H-labeled form Ic. Clearly, only form I DNA was labeled. Figure 4b shows, however, that during the chase in the presence of cycloheximide, only form Ic DNA was made. This was accompanied by the transfer of 32% of 4C-labeled form I DNA to the position of Ic. Dependency of the form I to Ic conversion on DNA replication. There are several possible explanations for the mechanism of conversion of form I DNA to form Ic. The question of whether an
DNA replication is involved in this process was examined as follows. Infected cells were labeled for 2.5 h with ["C ]TdR. This was followed by a 2-h pulse-label with ['H ]TdR in the presence of cycloheximide alone or in conjunction with cytosine arabinoside (ara-C). Primary isolation of viral DNA by velocity sedimentation in neutral sucrose gradients is shown in Fig. 5. "4C-labeled viral DNA synthesized in the absence of cycloheximide was comprised of form I DNA sedimenting at 20S, as expected (Fig. 5a). 'H-labeled DNA synthesized in the presence of cycloheximide sedimented in the position of form Ic (Fig. 5b; also see Fig. la). In addition, a shoulder of labeled material appeared on the trailing edge of the "4C-prelabeled viral DNA. In the presence of ara-C (Fig. 5c), ['H ]TdR incorporation into viral DNA was
VOL. 17, 1976
405
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
12 I0
;o x
8
15
x c
E
E
6
N
c
10
0
C
a 0
4 u
on
2
-
0 La 40
60
80
100
120
FRACT 0N FIG. 2.
60
40
80
100
120
NO.
Representative cesium chloride-PDIgradients of pulse-labeledform IDNA chased at of cycloheximide. Infected cells were labeled with [8H]TdR (150
in the presence
or
absence ACi/ml) 30.75 to 32.75 h postinfection. Radioactive medium was then replaced with medium containing unlabeled TdR with or without cycloheximide, and incubation was continued for 4 h. Viral DNA was isolated by neutral sucrose gradient sedimentation as described in Fig. la and analyzed by cesium chloride-PDI isopycnic centrifugation (rotor SW40). (a) Chase in the absence of cycloheximide. (b) Chase in the presence of cycloheximide. Symbols: 0, [3HJDNA; 0, "C-labeled form I DNA marker.
drastically reduced, and the shoulder on the "C-labeled form I DNA peak was not apparent. The residual [3HJDNA amounted to only 5% of the synthesis that occurred under the influence of cycloheximide but in the absence of ara-C. This material sedimented heterogeneously through the gradient rather than in an obvious peak. Analysis by cesium chloride-PDI equilibrium centrifugation showed that this DNA was almost entirely cellular in origin (data not shown). The indicated fractions in Fig. 5 were combined, and the DNA was concentrated by ethanol precipitation and analyzed in cesium chloride-PDI gradients. Results are shown in Fig. 6. "C-labeled form I DNA synthesized in the absence of cycloheximide (Fig. 6a) was partially converted to the position of Ic when cycloheximide was administered under conditions that allowed only the synthesis of Ic (Fig. 6b). However, no "C-labeled form Ic DNA was apparent in the presence of ara-C (Fig. 6c). These results indicate clearly that the formation of component Ic DNA from pre-existing form I templates depends upon DNA replication.
Replicative intermediate of form Ic DNA. Having established that the formation of component Ic occurs by a process dependent upon DNA replication, two alternative possibilities remained to explain the actual involvement of replication. We had previously shown that the amount of closed-circular viral DNA formed in the absence of protein synthesis is limited only by the rate of initiation of new rounds of genome replication (22). Thus, once the synthesis of form I DNA is initiated, replication proceeds normally, but some protein synthesis-dependent process in the closure of daughter DNA may result in reduction of superhelicity of the progeny molecules. On the other hand, it was also considered possible that the normal completion of a round of replication in the absence of protein synthesis would yield progeny DNA of normal superhelix density. Superhelical turns could then be removed from these molecules by a process independent of replication. To differentiate between these alternatives, we exammined the possibility that form I DNA is synthesized as an intermediate in the form I to Ic conversion.
Infected cells actively synthesizing viral DNA
J. VIROL.
YU AND CHEEVERS
406 0
(a)
8
C
6 ,*
0
0
4
c :3
2
v
0
(b)
u 40
30 i
* ..
0
.
0 0~~~~
3
4
t
20 10 0 .
..
0
2
CHASE (HOURS) FIG. 3. Kinetics of conversion of prelabeled form I DNA to form Ic in the absence of protein synthesis. (a) Inhibition of ['H]TdR incorporation into DNA by addition of excess unlabeled TdR (2 x 10-' M). (b) Accumulation of form Ic DNA from prelabeled form I in cycloheximide-treated cells. Data were derived according to procedures described in Fig. 2. Symbols: 0, Chase in the presence of cycloheximide; 0, chase in the absence of cycloheximide.
treated for 90 min with cycloheximide and then pulse-labeled for 4 min with ['H]TdR. Velocity sedimentation at neutral pH revealed labeled material with the sedimentation properties of viral replicative intermediate DNA (RI) (8) and form Ic (Fig. 7a). Cesium chloride-PDI analysis of this material (Fig. 7c) confirmed that a small amount of form Ic was generated during the pulse, but most of the DNA exhibited banding properties typical of viral RI (14). No detectable form I DNA was synthesized. In addition, viral RI chased in the absence of protein synthesis was entirely converted into closed-circular DNA of the form Ic conformation (Fig. 7b and d). Several experiments of the type shown in Fig. 7 have been done, varying the duration of the pulse label and the chase periods, and in no case has any form I DNA been detected. These results indicate that progeny form I DNA is not involved as an intermediate in the formation of component Ic. This evidence supports the hypothesis that form Ic arises via an alteration of the closure of newly replicated DNA rather than by removal of superhelical turns by a mechanism independent of replication. To verify this view, advantage was taken of the fact that the transition of form I DNA synthesis to form Ic in cycloheximide-treated cells occurs with first-order exponential kinetics (21), which allows a choice of were
conditions of treatment with cycloheximide in which both I and Ic DNA would be formed in predictable proportions. Thus the question was asked whether viral RI could be converted into a mixture of I and Ic, the proportions of which would remain stable during a long chase in the continued absence of protein synthesis. Cells were treated with cycloheximide under conditions of infection in which form Ic would be expected to represent about one-half of the closed-circular DNA being synthesized. The cultures were pulse-labeled for 3 min with ['H]TdR, and the newly synthesized DNA was then followed for an additional 2-h chase in the absence of protein synthesis. Figure 8a shows that the sedimentation distribution of the pulse-labeled DNA was that of viral RI. During the chase in the presence of cycloheximide, this DNA was converted into material with the sedimentation properties of a mixture of forms I and Ic closed-circular species. Figure 8b confirms by cesium chloride-PDI analysis that the closed-circular viral DNA present after the chase was comprised of about 40% form Ic and 60% form I. This experiment eliminates the possibilitythat superhelical turns may be removed from progeny DNA by a process independent of DNA replication. This is evident because the form I progeny DNA which resulted from the completion of viral RI labeled during the 3-min pulse was not converted to form Ic during the additional 2 h in the presence of cycloheximide. Thus, we conclude that protein synthesis is required for the normal closure of newly replicated polyoma DNA. After inhibition of protein synthesis by cycloheximide, some aspect of the closure mechanism is altered, resulting in the formation of progeny molecules exhibiting a marked deficiency in superhelicity. Metabolic fate of form Ic DNA. The fate of form Ic DNA was examined in pulse-chase experiments in which protein synthesis was allowed to resume during the chase period. Infected cells were pretreated with cycloheximide and incubated with [3H]TdR in the presence of cycloheximide to label newly synthesized Ic. The cultures were then washed with warm medium without cycloheximide, and incubation was continued with this medium for up to 4 h. Under these conditions, both protein synthesis and DNA synthesis returned to normal rates within less than 5 min after the removal of cycloheximide (data not shown). Figure 9 shows the distribution of labeled viral DNA in cesium chloride-PDI before and after chasing in the absence of cycloheximide. As expected, all of the DNA synthesized in the
3 0
0
x
x
01 x
E
E
2 .'
E
c
0
C
0
0
u I
u
I
VI
WI
WI
o
0
20
40
60
80
M
0
40
60
120
100
80
NO. F RACTION FIG. 4. Transfer of 'IC-labeled form I DNA into form Ic after a cycloheximide-induced block of protein synthesis. Infected cells were labeled with [14C]TdR (2 jACi/ml) at 30.75 to 32.75 h postinfection. Radioactive medium was then replaced with medium containing 10 ;Lg of cycloheximide per ml. Incubation was continued for 30 min to inhibit protein synthesis and then [3H]TdR was added to a final concentration of 40 MCi/ml. Viral DIVA was isolated as described in Fig. la and analyzed by cesium chloride-PDI isopycnic centrifugation as described in Fig. 2. (a) [14CJDNA (0), synthesized prior to the addition of cycloheximide, centrifuged with 'H-labeled form Ic marker DNA (0). (b) DNA synthesized in the presence of cycloheximide. Symbols: *, 'H-labeled radioactivity; 0, "C-labeled radioactivity. (b)
(a)
(C)
20s
4~~~~~~4' 20s
20s
''0
12I
a
C~~~~~~~~~~~~~~~~~~~
0~~~~~~~~~~~~~~~~~~~
3
5 10 152
5O51
01
52 2
2~~~~~~~~~~~~~~~~~
02 0
0
0
5
10
15 20 250
5
10
152025
0
5
10 1520 25 0
NO. F RACTION FIG. 5. Effect of ara-C on the generation of form Ic viral DNA. Velocity sedimentation profiles in neutral sucrose gradients of: (a) infected cells labeled with [14C]TdR (3 MCi/ml) at 30 to 32.5 h postinfection. (b) "4C-labeled viral DNA (as in a) followed by a 2-h labeling period with [3H]TdR (70,MCi/ml) in the presence of 10 Mg of cycloheximide per ml. (c) 14C-labeled viral DNA (as in a) followed by a 2-h labeling period with ['H]TdR (70 MCi/mI) in the presence of cycloheximide and 30 Mig of ara-C per ml. (0) 14C-labeled radioactivity; (0) 'H-labeled radioactivity. 407
408
J. VIROL.
YU AND CHEEVERS
16 14
12
10 8 6 4 2 0 N
0
200
N
14
C C*
E
0 x
150
I
12
c
I0
E
8
.-
6
0
4
0
4ft
100
C
0
z
0
50
In
2
0
0I 20
40
60
80
F RACT I0 N
100
120
NO.
FIG. 6. Cesium chloride-PDI analysis of viral DNA formed in the presence and absence of ara-C. DNA prepared by ethanol precipitation from the neutral sucrose gradient fractions indicated in Fig. 5 was analyzed by cesium chloride-PDI isopycnic centrifugation as described in Fig. 2. (a) ['4C]DNA from Fig. 5a centrifuged with 'H-labeled form Ic marker DNA. (b) DNA from Fig. 5b. (c) DNA from Fig. 5c centrifuged with 'H-labeled form Ic marker DNA. (0) 'H-labeled radioactivity; (0) "4C-labeled radioactivity.
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
VOL. 17, 1976
409
m
0 x c
E NC c U) 0
u x in
C 0
4
x
in C
C c
4-
0
0
u
u I
40
60
80
100
30 40
60
80
100
FRACTION NO. FIG. 7. Metabolic fate of viral RI in the absence ofprotein synthesis. Infected cells were pretreated for 90 min with cycloheximide at 28 h postinfection and pulse labeled with ['H]TdR (170 p Ci/mI) for 4 min. The radioactive medium was then replaced with medium containing 2 x 10-' M unlabeled TdR and cycloheximide. Viral DNA was isolated before (a) and after (b) the chase period by velocity sedimentation in neutral sucrose gradients. The indicated fractions were combined, and the DNA was concentrated by ethanol precipitation and analyzed by isopycnic centrifugation in cesium chloride-PDI as described in Fig. 2: (c) DNA from (a); (d) DNA from (b). (0) 'H-labeled radioactivity; (0) I'C-labeled form I marker DNA.
of cycloheximide banded in the position of form Ic (Fig. 9a). After a 2-h chase, upon restoration of protein and DNA synthesis, both form I and Ic were apparent, form Ic accounting for almost half of the DNA (Fig. 9b). After a 4-h chase (Fig. 9c), form Ic was reduced to one-third of the DNA, the remainder having been converted to form I. Figure 10b shows the kinetics of reacquisition by form Ic of the form I conformation upon reversal of cycloheximide inhibition. About half of the form Ic was converted to form I at an approximately linear rate within 90 min. Thereafter, conversion was markedly slower; by 4 h, 30 to 40% of the DNA still remained in the form Ic conformation. One experiment (not shown) indicated that this 30 to 40% residual Ic is stable for at least 10 h. The involvement of replication in the Ic to I presence
transition was examined by experiments in which form Ic was made in the presence of cycloheximide, and the labeled DNA was chased under conditions of restored protein synthesis in the presence or absence of hydroxyurea or ara-C to inhibit DNA synthesis. Figure 11 shows the cesium chloride-PDI density gradients obtained from a typical experiment: 60% of form Ic made in the absence of protein synthesis (Fig. 11a) was converted into form I DNA during a 4-h chase after reversal of cycloheximide inhibition (Fig. lib), The presence of hydroxyurea during the chase period had no effect on the proportions of I and Ic formed (Fig. lic). Similar results were obtained using ara-C to inhibit DNA synthesis. These results show that superhelical turns may be introduced into form Ic DNA upon restoration of protein synthesis by some process
410
J. VIROL.
YU AND CHEEVERS
(a) (0)
20s
20s 6
so
X
2
C
4 E
c
IN in
EIn
0
C
0
I:
0
I
2
n %J
10
10
15~~20 15 20
25
25
3
30
0
4 N
b 3
12 X c
10 E
E
c
if)
E
2
ro 4 2 0
0
0 10 20 30 40 50 60 70 FIG. 8. Metabolic stability of form I DNA in the absence of protein synthesis. (a) Velocity sedimentation analyses of viral DNA. Infected cells were pretreated for 30 min with cycloheximide and pulse labeled with [8HJTdR (150 ;Ci/ml) for 3 min. Cultures were then harvested (0) or chased for an additional 2 h with medium containing unlabeled TdR and cycloheximide (0). (A) "4C-labeled form I marker DNA. (b) DNA isolated from cultures after the chase period (indicated in a) was analyzed by cesium chloride-PDI centrifugation as described in Fig. 1. (0) [3HJDNA; (0) "4C-labeled form I marker DNA.
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
VOL. 17, 1976
411 12
10
on
0
9 C
x
8
E
7
-
6
c
E -
0
5
0
0
4
x U,
uo u
3
2 o
' 0
50 60
80
100
40
60
80
100
50 60
80
100
FRACTION NO. FIG. 9. Metabolic fate of form Ic DNA after removal of cycloheximide. Infected cells were pretreated with cycloheximide for 90 min and then labeled with ['H]TdR (40 uCi/ml) between 30.5 and 31.5 h postinfection in the continued presence of cycloheximide. Cultures were then harvested or chased with medium containing unlabeled TdR without cycloheximide for 1, 2, 3, or 4 h. Viral DNA was isolated as described in Fig. 1 and analyzed by cesium chloride-PDI centrifugation as described in Fig. 2. (a) One-hour pulse with [3H]TdR in cycloheximide-treated cells; (b) 1-h pulse with ['H]TdR in the presence of cycloheximide followed by 2-h chase in the absence of cycloheximide; (c) 1-h pulse with [3H]TdR in the presence of cycloheximide followed by 4-h chase in the absence of cycloheximide. (0) [3H]DNA; (0) "C-labeled form I DNA marker.
independent of DNA replication. A subpopula- monomers made in the presence of cyclohexition of Ic molecules, accounting for about one- mide. third of the total, does not respond to cyclohexiExperiments in which newly replicated form I mide reversal by reacquisition of the normal DNA was chased in cycloheximide-treated cells superhelix density of polyoma DNA. These indicate that form Ic arises from pre-existing molecules are apparently stable end products of form I molecules. Approximately one-third of replication in the absence of protein synthesis. the form I DNA replicated during a 2-h interval are converted to form Ic during a subsequent DISCUSSION 2-h chase in the absence of protein synthesis. This study confirms our previous work (6) This proportion is not significantly changed by and that of Bourgaux and Bourgaux-Ramoisy increasing the chase time. The point to be made (2), indicating that monomeric viral DNA with from this experiment is that only one-third of a deficiency in superhelicity is synthesized in the form I molecules can be traced into the Ic polyoma virus-infected cells in which protein population at a time when essentially 100% of synthesis has been blocked. We have termed viral DNA being replicated is accounted for by this new DNA species form Ic. The proportion of the formation of Ic. The most obvious explanation of this finding oligomeric viral DNA is also increased in the absence of protein synthesis, and these mole- assumes that form Ic arises by replication on cules also exhibit a reduction in superhelical form I templates, since the rate of initiation of turns (1; Yu and Cheevers, unpublished data). new rounds of viral genome replication is inhibSimilar effects have also been noted in simian ited by cycloheximide (22) by an amount convirus 40-infected cells for both monomers (20) sistent with the proportion of form I DNA and oligomers (12). The present work was begun involved in the formation of Ic (21). The into determine the mechanism of formation of volvement of replication in the conversion of component Ic closed-circular polyoma DNA form I to Ic was confirmed by the sensitivity of
J. VIROL.
YU AND CHEEVERS
412
was not supported, as no evidence was found to suggest that superhelical turns are removed from newly synthesized DNA by a mechanism
6 x E
4
XC)
fo
A
A
11
)
A
2
0
,(b)
100
0
90
80
70 Q6 60 0
\~~~
5 D
-
4 0
30 20 10
0 0
2
4
Chase (hours)
FIG. 10. Kinetics of formation of forr I DNA from prelabeled form Ic upon restoration of p rotein synthesis. (a) Inhibition of [3H]TdR incorporatetiol into DNA by addition of excess unlabeled TdR (2 10-DM). (b) Kinetics of reformation of component removal of cycloheximide. Data were derived from three separate experiments as described A) Form Ic DNA chased in the presencee of cycloheximide; (S, A, *) form Ic DNA chased in the absence of cycloheximide. n
x
indFig. (o
independent of replication. In fact, the data indicate that forms I and Ic arise from a common replicative intermediate population. Inhibition of protein synthesis gradually shifts the synthesis of viral DNA from form I to Ic (21). Under conditions of treatment with cycloheximide in which viral RI is chased into a mixture of closed-circular progeny molecules, the proportions of forms I and Ic are stable in the continued absence of protein synthesis. Thus, we conclude that the formation of component Ic viral DNA results from an alteration of some protein synthesis-dependent process in the closure of daughter molecules. Upon restoration of protein synthesis, approximately two-thirds of form Ic molecules return to the form I conformation by a process independent of DNA synthesis. This result supports previous work with simian virus 40 indicating that supercoiling of viral DNA does not depend upon replication (20). The failure of
one-third of the form Ic molecules to assume a normal superhelix density upon restoration of protein synthesis is not understood. This is not due, however, to incomplete reversal of the effect of cycloheximide. These molecules may be substituted with cellular DNA sequences,
arising by cycloheximide-induced enhancement of excision of integrated viral DNA, since highthis process to the inhibition of Db JA synthesis multiplicity passage yields such DNA molecules that band at the position of form Ic in cesium by ara-C. Two alternative mechanisms wer e considered chloride-PDI (Hiscott, Yu, and Cheevers, unto account for the synthesis of for]m Ic. It was published data). This possibility is being invesassumed that form I DNA mole,cules whose tigated. Figure 12 shows a diagrammatic summary of synthesis had already been initiate(d at the time of addition of cycloheximide as Mvell as those the results of this work, which assumes that initiated after the inhibition of prc)tein synthe- form II is a terminal intermediate of polyoma sis would finish replication normall' y (22). Thus, DNA replication (10). Upon inhibition of proform Ic could arise either by some alteration of tein synthesis, the synthesis of polyoma DNA is the closure of daughter molecules o r by removal limited by a reduced rate of initiation of new of superhelical turns from newly synthesized rounds of genome replication (22; step 1). Form form I DNA. The latter possibiility seemed I DNA molecules initiated in the absence of particularly interesting in view of t-he discovery protein synthesis finish replication normally, by Wang (19) of protein in Esct Lerichia coli, yielding closed-circular progeny DNA (22). Prowhich removes superhelical turns iin closed-cir- tein synthesis is also required to maintain the cular DNA without introducing permanent superhelix density of progeny DNA. In its strand scissions. Similar "untwistinig" activities absence (step 2) the synthesis of viral DNA have been found in extracts of m Duse (5) and shifts from the formation of normal form I human cells (13), as well as in nlucleoprotein progeny to form Ic, a monomeric closed-circular complexes containing replicating
JOURNAL OF VIROLOGY, Feb. 1976, p. 402-414 Copyright 0 1976 American Society for Microbiology
DNA Synthesis in Polyoma Virus Infection IV. Mechanism of Formation of Closed-Circular Viral DNA Deficient in Superhelical Turns KENNETH YU AND WILLIAM P. CHEEVERS* Cancer Research Laboratory, University of Western Ontario, London, Ontario, Canada Received for publication 22 July 1975
A marked reduction in the rate of viral DNA synthesis is accompanied by an alteration to the superhelicity of progeny DNA in polyoma virus-infected cells in which protein synthesis has been inhibited by cycloheximide. Viral DNA molecules formed in the presence of cycloheximide consist predominantly of closed-circular monomeric species (referred to as form Ic) characterized by a decreased superhelix density, corresponding to Aao, = 0.0195, as compared to form I DNA by propidium diiodide-cesium chloride isopycnic analysis. Form Ic is synthesized on pre-existing form I templates without the intervention of progeny form I as an intermediate. It is concluded that inhibition of protein synthesis results in the alteration of some process in the closure of daughter DNA that leads to a marked reduction of superhelical turns of progeny molecules. About two-thirds of form Ic molecules return to the form I conformation upon reversal of cycloheximide inhibition by a mechanism independent of DNA replication.
When protein synthesis is inhibited in polyoma virus-infected cells by either cycloheximide (3, 6) or puromycin (2), a viral DNA component is formed that sediments at neutral pH slightly slower than 20S form I viral DNA. Alkaline velocity sedimentation and isopycnic centrifugation in propidium diiodide-cesium chloride show this DNA (form Ic) to consist of monomeric closed-circular molecules with a decreased superhelicity as compared to form I (2, 6). The present study is concerned with the mechanism of formation of Ic viral DNA in cycloheximide-treated cells. It is shown that newly synthesized Ic is made on pre-existing form I DNA templates. Closed-circular progeny DNA of normal superhelix density is not involved as an intermediate in this process. It is concluded that form Ic arises by a cycloheximide-induced alteration of the closure of daughter molecules. Improperly closed progeny DNA may reacquire the normal tertiary structure of form I upon restoration of protein synthesis. This process does not involve DNA replication. (This work was taken from a thesis by K. Y. for submission to the Department of Bacteriology and Immunology, University of Western Ontario, in partial fulfillment of requirements for the Ph.D. degree.) MATERIALS AND METHODS Tissue culture and virus infection. Primary cul402
tures of mouse embryo cells were grown in McCoy 5A medium supplemented with 10% fetal calf serum at 37 C in 5% CO2-95% air. For experiments, primary cells were subcultured into 100-mm Falcon plastic dishes at 15 x 101 cells per dish. Cultures were grown to stationary phase (approximately 40 x 105 cells) and infected with TSP1 polyoma virus (17) at a multiplicity of 30 plaque-forming units per cell as previously described (4). Isotopic labeling and extraction of DNA. Replicating viral DNA was radioactively labeled by exposure of polyoma-infected cultures to medium containing [methyl-3H ]thymidine (TdR) (New England Nuclear; 40 to 52 Ci/mmol) or [2- "C]TdR (Amersham/ Searle; 40 to 60 mCi/mmol). The conditions of labeling are given in individual experiments. For pulsechase experiments, the labeled medium was decanted and the cells were immediately washed twice with prewarmed (37 C) medium containing 2 x 10-i M TdR and then incubated with this medium for the required chase time. For extraction of DNA, cells were rinsed with ice-cold SSC (0.15 M NaCl plus 0.015 M trisodium citrate) and dispersed by 0.1% trypsin in saline citrate solution (16). Cells were then collected by centrifugation, suspended in SSC and lysed by addition of 0.1 volume of 10% sodium dodecyl sulfate in EDTA buffer (0.1 M NaCl-0.001 M EDTA-0.01 M Tris, pH 7.4). Lysates were centrifuged through 30-ml 15 to 30% (wt/wt) sucrose gradients in EDTA buffer containing 0.5% sodium dodecyl sulfate, formed over a 6-ml cushion of 70% (wt/vol) sucrose in EDTA buffer. Centrifugation was at 26,000 rpm at 23 C for 10 h using a Spinco SW27 rotor. Under these conditions approximately 95% of the cellular DNA in the crude lysates is recovered as high-molecular-weight mate-
VOL. 17, 1976
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
403
cycloheximide. Two points are evident: (i) much less viral DNA was synthesized than in untreated cultures (this effect has been shown to result exclusively from limited initiation of new rounds of genome replication [21 ]); and (ii) 20S form I DNA was replaced by a DNA component sedimenting at 16S. This component is form Ic. Form Ic sediments as monomeric closed-circular viral DNA under alkaline conditions (6) and shows a decreased superhelix density by isopycnic analysis in cesium chloride-PDI (Fig. ic) as compared to form I DNA (Fig. lb). The superhelix density of form Ic was determined as follows. Forms I and Ic, labeled with [14C ]TdR and [3H]TdR, respectively, were isolated as described in Fig. la and purified by cesium chloride-PDI centrifugation followed by Sephadex G-50 column chromatography to remove CsCl and PDI. After concentration by ethanol precipitation, purified forms I and Ic were mixed, partially converted to form H by limited digestion with deoxyribonuclease, and centrifuged to equilibrium in cesium chloridePDI. The distance between the bands corresponding to 3H-labeled II and 3H-labeled Ic was compared with that of "4C-labeled II and 14C_ labeled I according to the Eason and Vinograd equation (9) Aa. = 0.1 [(Ar/Ar*) - 1]. Our measurements indicate a shift in superhelix density from I to Ic equivalent to &ao = 0.0195, which corresponds to a decrease in the number of superhelical turns of about two-thirds (13, 18). Origin of form Ic DNA. The purpose of this experiment was to test the possibility that form Ic arises from form I molecules. Replicating form I DNA was labeled for 2 h with [3H]TdR. The fate of these labeled molecules was then determined during a chase for various periods of time with unlabeled TdR in the presence or absence of cycloheximide. Figure 2 shows the cesium chloride-PDI density gradient profiles of viral DNA chased for 4 h. Without cycloheximide (Fig. 2a), [3HIDNA banded with marker "4C-labeled form I as expected. In the absence of protein synthesis, however, a portion of 3H-labeled form I DNA was chased into the position of Ic (Fig. 2b). The quantitative transfer of label from I to Ic during the 4-h chase period in the presence of cycloheximide is shown in Fig. 3b. These data show that about one-third of form I DNA molecules replicated during a 2-h period were diverted to the RESULTS formation of component Ic during a subsequent Characterization of form Ic DNA. Figure la 2-h period in the absence of protein synthesis. This finding was confirmed by the experishows the sedimentation properties at neutral pH of polyoma DNA formed in the presence of ment described in Fig. 4. Infected cells were
rial on the dense sucrose cushion, whereas polyoma DNA components are resolved within the gradient (6-8, 22). Viral DNA was recovered for subsequent analysis as follows. Appropriate neutral sucrose gradient fractions were combined, and the DNA was precipitated with 2.5 volumes of cold 95% ethanol. After overnight storage at -20 C, the precipitate was collected by centrifugation (10,000 x g; 30 min; 0 C) and dissolved in 0.3 M NaCl-0.001 M EDTA-0.01 M Tris, pH 8.1. Dye-buoyant density centrifugation. The superhelix density of closed-circular polyoma DNA was analyzed by equilibrium centrifugation in CsCl gradients containing propidium diiodide (PDI) (11). i4Clabeled form I marker DNA was prepared by labeling infected cultures with ["4C]TdR (0.2 to 0.5 gCi/ml) from 10 to 30 h postinfection. 'H-labeled form Ic marker DNA was prepared from infected cultures labeled with [3H ]TdR (50 IACi/ml) at 28 to 34 h postinfection in the presence of cycloheximide. Test DNA samples were mixed with the appropriate marker and a solution of CsCl in 0.02 M Tris (pH 8.6)-0.02 M EDTA containing PDI to give a final volume of 3.5 ml, an initial density of 1.52 g/cm3, and a final concentration of 500 gg of PDI/ml. These solutions were then centrifuged to equilibrium at 20 C in a Spinco SW40 rotor for 90 h at 22,000 rpm or alternatively in the SW50L or SW50.1 rotor for 48 h at 40,000 rpm. Radioactivity determinations. For measurement of total radioactive DNA, cell cultures were lysed by addition of 0.5% sodium dodecyl sulfate in EDTA buffer. Lysates were then mixed with an equal volume of cold 10% trichloroacetic acid to precipitate the macromolecular fraction. Acid-insoluble material was collected onto Whatman GF/C glass fiber filters and counted in nonaqueous scintillator as previously described (8). Sucrose gradients were collected using an ISCO model 640 fractionator equipped with model UA-2 UV analyzer at 254 nm. Cesium chloride-PDI gradients were collected using a Buchler Instruments fraction collector. Three-drop fractions were collected using constant positive pressure from the top generated by a multispeed transmission (Harvard Apparatus Co., Dover, Mass.) pushing a column of paraffin oil at the rate of 0.408 ml/min. Fractions were mixed with an excess of cold 5% trichloroacetic acid and the acidinsoluble fraction was collected onto membrane (Millipore) or cellulose ester (Gelman) filters (0.45-um pore size). The filters were washed with 5% trichloroacetic acid in 95% ethanol and counted as described above. Alternatively, fractions were mixed with two drops of an aqueous solution of fraction V bovine serum albumin (1 mg/ml) and adjusted to a final concentration of 5% trichloroacetic acid, and the acid-insoluble fraction was collected by filtration onto Whatman GF/C glass fiber papers.
404
J. VIROL.
YU AND CHEEVERS
c 0 a-
on
x
0
tvb
a
x
E
"._
In.
e0
go
c
0 =x 0
-
0-
N.
C
C
0
E
u
0
U
N1-.
lu e
on
0-
0
u
0
10
20
30
40
FRACTION NO.
FRACTION NO.
FIG. 1. Sedimentation and dye-buoyant density centrifugation properties of polyoma DNA synthesized in cycloheximide-treated cells. (a) Velocity sedimentation at neutral pH. Infected cells were pretreated for 90 min at 28 h postinfection with medium with or without cycloheximide (10 jug/ml) and then pulse labeled for I h with ['H]TdR (25 jCi/ml). Cell lysates were prepared by treatment with sodium dodecyl sulfate and centrifuged in neutral sucrose gradients as described. Symbols: 0, Untreated cells; 0, cycloheximide-treated cells. The direction of sedimentation is from left to right in all velocity sedimentation analyses. The indicated fractions in (a) were combined and the DNA was concentrated by ethanol precipitation and analyzed by isopycnic centrifugation in cesium chloride-PDI (rotor SW50.1). (b) Untreated cells. Symbols: 0, ['HJDNA; x, "C-labeled marker form IDNA. (c) Cycloheximide-treated cells. Symbols: 0, [3HIDNA; x, "4C-labeled marker form IDNA. Density increases from right to left in all cesium chloride-PDI gradients.
labeled for 2 h with ["C ]TdR and then chased additional 2 h with medium containing unlabeled TdR and cycloheximide. ['H ]TdR was administered during the last 1.5 h of the chase period to label newly formed component Ic. Figure 4a shows the distribution in cesium chloride-PDI of closed-circular "C-labeled viral DNA before the chase period, relative to that of marker 'H-labeled form Ic. Clearly, only form I DNA was labeled. Figure 4b shows, however, that during the chase in the presence of cycloheximide, only form Ic DNA was made. This was accompanied by the transfer of 32% of 4C-labeled form I DNA to the position of Ic. Dependency of the form I to Ic conversion on DNA replication. There are several possible explanations for the mechanism of conversion of form I DNA to form Ic. The question of whether an
DNA replication is involved in this process was examined as follows. Infected cells were labeled for 2.5 h with ["C ]TdR. This was followed by a 2-h pulse-label with ['H ]TdR in the presence of cycloheximide alone or in conjunction with cytosine arabinoside (ara-C). Primary isolation of viral DNA by velocity sedimentation in neutral sucrose gradients is shown in Fig. 5. "4C-labeled viral DNA synthesized in the absence of cycloheximide was comprised of form I DNA sedimenting at 20S, as expected (Fig. 5a). 'H-labeled DNA synthesized in the presence of cycloheximide sedimented in the position of form Ic (Fig. 5b; also see Fig. la). In addition, a shoulder of labeled material appeared on the trailing edge of the "4C-prelabeled viral DNA. In the presence of ara-C (Fig. 5c), ['H ]TdR incorporation into viral DNA was
VOL. 17, 1976
405
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
12 I0
;o x
8
15
x c
E
E
6
N
c
10
0
C
a 0
4 u
on
2
-
0 La 40
60
80
100
120
FRACT 0N FIG. 2.
60
40
80
100
120
NO.
Representative cesium chloride-PDIgradients of pulse-labeledform IDNA chased at of cycloheximide. Infected cells were labeled with [8H]TdR (150
in the presence
or
absence ACi/ml) 30.75 to 32.75 h postinfection. Radioactive medium was then replaced with medium containing unlabeled TdR with or without cycloheximide, and incubation was continued for 4 h. Viral DNA was isolated by neutral sucrose gradient sedimentation as described in Fig. la and analyzed by cesium chloride-PDI isopycnic centrifugation (rotor SW40). (a) Chase in the absence of cycloheximide. (b) Chase in the presence of cycloheximide. Symbols: 0, [3HJDNA; 0, "C-labeled form I DNA marker.
drastically reduced, and the shoulder on the "C-labeled form I DNA peak was not apparent. The residual [3HJDNA amounted to only 5% of the synthesis that occurred under the influence of cycloheximide but in the absence of ara-C. This material sedimented heterogeneously through the gradient rather than in an obvious peak. Analysis by cesium chloride-PDI equilibrium centrifugation showed that this DNA was almost entirely cellular in origin (data not shown). The indicated fractions in Fig. 5 were combined, and the DNA was concentrated by ethanol precipitation and analyzed in cesium chloride-PDI gradients. Results are shown in Fig. 6. "C-labeled form I DNA synthesized in the absence of cycloheximide (Fig. 6a) was partially converted to the position of Ic when cycloheximide was administered under conditions that allowed only the synthesis of Ic (Fig. 6b). However, no "C-labeled form Ic DNA was apparent in the presence of ara-C (Fig. 6c). These results indicate clearly that the formation of component Ic DNA from pre-existing form I templates depends upon DNA replication.
Replicative intermediate of form Ic DNA. Having established that the formation of component Ic occurs by a process dependent upon DNA replication, two alternative possibilities remained to explain the actual involvement of replication. We had previously shown that the amount of closed-circular viral DNA formed in the absence of protein synthesis is limited only by the rate of initiation of new rounds of genome replication (22). Thus, once the synthesis of form I DNA is initiated, replication proceeds normally, but some protein synthesis-dependent process in the closure of daughter DNA may result in reduction of superhelicity of the progeny molecules. On the other hand, it was also considered possible that the normal completion of a round of replication in the absence of protein synthesis would yield progeny DNA of normal superhelix density. Superhelical turns could then be removed from these molecules by a process independent of replication. To differentiate between these alternatives, we exammined the possibility that form I DNA is synthesized as an intermediate in the form I to Ic conversion.
Infected cells actively synthesizing viral DNA
J. VIROL.
YU AND CHEEVERS
406 0
(a)
8
C
6 ,*
0
0
4
c :3
2
v
0
(b)
u 40
30 i
* ..
0
.
0 0~~~~
3
4
t
20 10 0 .
..
0
2
CHASE (HOURS) FIG. 3. Kinetics of conversion of prelabeled form I DNA to form Ic in the absence of protein synthesis. (a) Inhibition of ['H]TdR incorporation into DNA by addition of excess unlabeled TdR (2 x 10-' M). (b) Accumulation of form Ic DNA from prelabeled form I in cycloheximide-treated cells. Data were derived according to procedures described in Fig. 2. Symbols: 0, Chase in the presence of cycloheximide; 0, chase in the absence of cycloheximide.
treated for 90 min with cycloheximide and then pulse-labeled for 4 min with ['H]TdR. Velocity sedimentation at neutral pH revealed labeled material with the sedimentation properties of viral replicative intermediate DNA (RI) (8) and form Ic (Fig. 7a). Cesium chloride-PDI analysis of this material (Fig. 7c) confirmed that a small amount of form Ic was generated during the pulse, but most of the DNA exhibited banding properties typical of viral RI (14). No detectable form I DNA was synthesized. In addition, viral RI chased in the absence of protein synthesis was entirely converted into closed-circular DNA of the form Ic conformation (Fig. 7b and d). Several experiments of the type shown in Fig. 7 have been done, varying the duration of the pulse label and the chase periods, and in no case has any form I DNA been detected. These results indicate that progeny form I DNA is not involved as an intermediate in the formation of component Ic. This evidence supports the hypothesis that form Ic arises via an alteration of the closure of newly replicated DNA rather than by removal of superhelical turns by a mechanism independent of replication. To verify this view, advantage was taken of the fact that the transition of form I DNA synthesis to form Ic in cycloheximide-treated cells occurs with first-order exponential kinetics (21), which allows a choice of were
conditions of treatment with cycloheximide in which both I and Ic DNA would be formed in predictable proportions. Thus the question was asked whether viral RI could be converted into a mixture of I and Ic, the proportions of which would remain stable during a long chase in the continued absence of protein synthesis. Cells were treated with cycloheximide under conditions of infection in which form Ic would be expected to represent about one-half of the closed-circular DNA being synthesized. The cultures were pulse-labeled for 3 min with ['H]TdR, and the newly synthesized DNA was then followed for an additional 2-h chase in the absence of protein synthesis. Figure 8a shows that the sedimentation distribution of the pulse-labeled DNA was that of viral RI. During the chase in the presence of cycloheximide, this DNA was converted into material with the sedimentation properties of a mixture of forms I and Ic closed-circular species. Figure 8b confirms by cesium chloride-PDI analysis that the closed-circular viral DNA present after the chase was comprised of about 40% form Ic and 60% form I. This experiment eliminates the possibilitythat superhelical turns may be removed from progeny DNA by a process independent of DNA replication. This is evident because the form I progeny DNA which resulted from the completion of viral RI labeled during the 3-min pulse was not converted to form Ic during the additional 2 h in the presence of cycloheximide. Thus, we conclude that protein synthesis is required for the normal closure of newly replicated polyoma DNA. After inhibition of protein synthesis by cycloheximide, some aspect of the closure mechanism is altered, resulting in the formation of progeny molecules exhibiting a marked deficiency in superhelicity. Metabolic fate of form Ic DNA. The fate of form Ic DNA was examined in pulse-chase experiments in which protein synthesis was allowed to resume during the chase period. Infected cells were pretreated with cycloheximide and incubated with [3H]TdR in the presence of cycloheximide to label newly synthesized Ic. The cultures were then washed with warm medium without cycloheximide, and incubation was continued with this medium for up to 4 h. Under these conditions, both protein synthesis and DNA synthesis returned to normal rates within less than 5 min after the removal of cycloheximide (data not shown). Figure 9 shows the distribution of labeled viral DNA in cesium chloride-PDI before and after chasing in the absence of cycloheximide. As expected, all of the DNA synthesized in the
3 0
0
x
x
01 x
E
E
2 .'
E
c
0
C
0
0
u I
u
I
VI
WI
WI
o
0
20
40
60
80
M
0
40
60
120
100
80
NO. F RACTION FIG. 4. Transfer of 'IC-labeled form I DNA into form Ic after a cycloheximide-induced block of protein synthesis. Infected cells were labeled with [14C]TdR (2 jACi/ml) at 30.75 to 32.75 h postinfection. Radioactive medium was then replaced with medium containing 10 ;Lg of cycloheximide per ml. Incubation was continued for 30 min to inhibit protein synthesis and then [3H]TdR was added to a final concentration of 40 MCi/ml. Viral DIVA was isolated as described in Fig. la and analyzed by cesium chloride-PDI isopycnic centrifugation as described in Fig. 2. (a) [14CJDNA (0), synthesized prior to the addition of cycloheximide, centrifuged with 'H-labeled form Ic marker DNA (0). (b) DNA synthesized in the presence of cycloheximide. Symbols: *, 'H-labeled radioactivity; 0, "C-labeled radioactivity. (b)
(a)
(C)
20s
4~~~~~~4' 20s
20s
''0
12I
a
C~~~~~~~~~~~~~~~~~~~
0~~~~~~~~~~~~~~~~~~~
3
5 10 152
5O51
01
52 2
2~~~~~~~~~~~~~~~~~
02 0
0
0
5
10
15 20 250
5
10
152025
0
5
10 1520 25 0
NO. F RACTION FIG. 5. Effect of ara-C on the generation of form Ic viral DNA. Velocity sedimentation profiles in neutral sucrose gradients of: (a) infected cells labeled with [14C]TdR (3 MCi/ml) at 30 to 32.5 h postinfection. (b) "4C-labeled viral DNA (as in a) followed by a 2-h labeling period with [3H]TdR (70,MCi/ml) in the presence of 10 Mg of cycloheximide per ml. (c) 14C-labeled viral DNA (as in a) followed by a 2-h labeling period with ['H]TdR (70 MCi/mI) in the presence of cycloheximide and 30 Mig of ara-C per ml. (0) 14C-labeled radioactivity; (0) 'H-labeled radioactivity. 407
408
J. VIROL.
YU AND CHEEVERS
16 14
12
10 8 6 4 2 0 N
0
200
N
14
C C*
E
0 x
150
I
12
c
I0
E
8
.-
6
0
4
0
4ft
100
C
0
z
0
50
In
2
0
0I 20
40
60
80
F RACT I0 N
100
120
NO.
FIG. 6. Cesium chloride-PDI analysis of viral DNA formed in the presence and absence of ara-C. DNA prepared by ethanol precipitation from the neutral sucrose gradient fractions indicated in Fig. 5 was analyzed by cesium chloride-PDI isopycnic centrifugation as described in Fig. 2. (a) ['4C]DNA from Fig. 5a centrifuged with 'H-labeled form Ic marker DNA. (b) DNA from Fig. 5b. (c) DNA from Fig. 5c centrifuged with 'H-labeled form Ic marker DNA. (0) 'H-labeled radioactivity; (0) "4C-labeled radioactivity.
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
VOL. 17, 1976
409
m
0 x c
E NC c U) 0
u x in
C 0
4
x
in C
C c
4-
0
0
u
u I
40
60
80
100
30 40
60
80
100
FRACTION NO. FIG. 7. Metabolic fate of viral RI in the absence ofprotein synthesis. Infected cells were pretreated for 90 min with cycloheximide at 28 h postinfection and pulse labeled with ['H]TdR (170 p Ci/mI) for 4 min. The radioactive medium was then replaced with medium containing 2 x 10-' M unlabeled TdR and cycloheximide. Viral DNA was isolated before (a) and after (b) the chase period by velocity sedimentation in neutral sucrose gradients. The indicated fractions were combined, and the DNA was concentrated by ethanol precipitation and analyzed by isopycnic centrifugation in cesium chloride-PDI as described in Fig. 2: (c) DNA from (a); (d) DNA from (b). (0) 'H-labeled radioactivity; (0) I'C-labeled form I marker DNA.
of cycloheximide banded in the position of form Ic (Fig. 9a). After a 2-h chase, upon restoration of protein and DNA synthesis, both form I and Ic were apparent, form Ic accounting for almost half of the DNA (Fig. 9b). After a 4-h chase (Fig. 9c), form Ic was reduced to one-third of the DNA, the remainder having been converted to form I. Figure 10b shows the kinetics of reacquisition by form Ic of the form I conformation upon reversal of cycloheximide inhibition. About half of the form Ic was converted to form I at an approximately linear rate within 90 min. Thereafter, conversion was markedly slower; by 4 h, 30 to 40% of the DNA still remained in the form Ic conformation. One experiment (not shown) indicated that this 30 to 40% residual Ic is stable for at least 10 h. The involvement of replication in the Ic to I presence
transition was examined by experiments in which form Ic was made in the presence of cycloheximide, and the labeled DNA was chased under conditions of restored protein synthesis in the presence or absence of hydroxyurea or ara-C to inhibit DNA synthesis. Figure 11 shows the cesium chloride-PDI density gradients obtained from a typical experiment: 60% of form Ic made in the absence of protein synthesis (Fig. 11a) was converted into form I DNA during a 4-h chase after reversal of cycloheximide inhibition (Fig. lib), The presence of hydroxyurea during the chase period had no effect on the proportions of I and Ic formed (Fig. lic). Similar results were obtained using ara-C to inhibit DNA synthesis. These results show that superhelical turns may be introduced into form Ic DNA upon restoration of protein synthesis by some process
410
J. VIROL.
YU AND CHEEVERS
(a) (0)
20s
20s 6
so
X
2
C
4 E
c
IN in
EIn
0
C
0
I:
0
I
2
n %J
10
10
15~~20 15 20
25
25
3
30
0
4 N
b 3
12 X c
10 E
E
c
if)
E
2
ro 4 2 0
0
0 10 20 30 40 50 60 70 FIG. 8. Metabolic stability of form I DNA in the absence of protein synthesis. (a) Velocity sedimentation analyses of viral DNA. Infected cells were pretreated for 30 min with cycloheximide and pulse labeled with [8HJTdR (150 ;Ci/ml) for 3 min. Cultures were then harvested (0) or chased for an additional 2 h with medium containing unlabeled TdR and cycloheximide (0). (A) "4C-labeled form I marker DNA. (b) DNA isolated from cultures after the chase period (indicated in a) was analyzed by cesium chloride-PDI centrifugation as described in Fig. 1. (0) [3HJDNA; (0) "4C-labeled form I marker DNA.
POLYOMA DNA SUPERHELICITY AND CYCLOHEXIMIDE
VOL. 17, 1976
411 12
10
on
0
9 C
x
8
E
7
-
6
c
E -
0
5
0
0
4
x U,
uo u
3
2 o
' 0
50 60
80
100
40
60
80
100
50 60
80
100
FRACTION NO. FIG. 9. Metabolic fate of form Ic DNA after removal of cycloheximide. Infected cells were pretreated with cycloheximide for 90 min and then labeled with ['H]TdR (40 uCi/ml) between 30.5 and 31.5 h postinfection in the continued presence of cycloheximide. Cultures were then harvested or chased with medium containing unlabeled TdR without cycloheximide for 1, 2, 3, or 4 h. Viral DNA was isolated as described in Fig. 1 and analyzed by cesium chloride-PDI centrifugation as described in Fig. 2. (a) One-hour pulse with [3H]TdR in cycloheximide-treated cells; (b) 1-h pulse with ['H]TdR in the presence of cycloheximide followed by 2-h chase in the absence of cycloheximide; (c) 1-h pulse with [3H]TdR in the presence of cycloheximide followed by 4-h chase in the absence of cycloheximide. (0) [3H]DNA; (0) "C-labeled form I DNA marker.
independent of DNA replication. A subpopula- monomers made in the presence of cyclohexition of Ic molecules, accounting for about one- mide. third of the total, does not respond to cyclohexiExperiments in which newly replicated form I mide reversal by reacquisition of the normal DNA was chased in cycloheximide-treated cells superhelix density of polyoma DNA. These indicate that form Ic arises from pre-existing molecules are apparently stable end products of form I molecules. Approximately one-third of replication in the absence of protein synthesis. the form I DNA replicated during a 2-h interval are converted to form Ic during a subsequent DISCUSSION 2-h chase in the absence of protein synthesis. This study confirms our previous work (6) This proportion is not significantly changed by and that of Bourgaux and Bourgaux-Ramoisy increasing the chase time. The point to be made (2), indicating that monomeric viral DNA with from this experiment is that only one-third of a deficiency in superhelicity is synthesized in the form I molecules can be traced into the Ic polyoma virus-infected cells in which protein population at a time when essentially 100% of synthesis has been blocked. We have termed viral DNA being replicated is accounted for by this new DNA species form Ic. The proportion of the formation of Ic. The most obvious explanation of this finding oligomeric viral DNA is also increased in the absence of protein synthesis, and these mole- assumes that form Ic arises by replication on cules also exhibit a reduction in superhelical form I templates, since the rate of initiation of turns (1; Yu and Cheevers, unpublished data). new rounds of viral genome replication is inhibSimilar effects have also been noted in simian ited by cycloheximide (22) by an amount convirus 40-infected cells for both monomers (20) sistent with the proportion of form I DNA and oligomers (12). The present work was begun involved in the formation of Ic (21). The into determine the mechanism of formation of volvement of replication in the conversion of component Ic closed-circular polyoma DNA form I to Ic was confirmed by the sensitivity of
J. VIROL.
YU AND CHEEVERS
412
was not supported, as no evidence was found to suggest that superhelical turns are removed from newly synthesized DNA by a mechanism
6 x E
4
XC)
fo
A
A
11
)
A
2
0
,(b)
100
0
90
80
70 Q6 60 0
\~~~
5 D
-
4 0
30 20 10
0 0
2
4
Chase (hours)
FIG. 10. Kinetics of formation of forr I DNA from prelabeled form Ic upon restoration of p rotein synthesis. (a) Inhibition of [3H]TdR incorporatetiol into DNA by addition of excess unlabeled TdR (2 10-DM). (b) Kinetics of reformation of component removal of cycloheximide. Data were derived from three separate experiments as described A) Form Ic DNA chased in the presencee of cycloheximide; (S, A, *) form Ic DNA chased in the absence of cycloheximide. n
x
indFig. (o
independent of replication. In fact, the data indicate that forms I and Ic arise from a common replicative intermediate population. Inhibition of protein synthesis gradually shifts the synthesis of viral DNA from form I to Ic (21). Under conditions of treatment with cycloheximide in which viral RI is chased into a mixture of closed-circular progeny molecules, the proportions of forms I and Ic are stable in the continued absence of protein synthesis. Thus, we conclude that the formation of component Ic viral DNA results from an alteration of some protein synthesis-dependent process in the closure of daughter molecules. Upon restoration of protein synthesis, approximately two-thirds of form Ic molecules return to the form I conformation by a process independent of DNA synthesis. This result supports previous work with simian virus 40 indicating that supercoiling of viral DNA does not depend upon replication (20). The failure of
one-third of the form Ic molecules to assume a normal superhelix density upon restoration of protein synthesis is not understood. This is not due, however, to incomplete reversal of the effect of cycloheximide. These molecules may be substituted with cellular DNA sequences,
arising by cycloheximide-induced enhancement of excision of integrated viral DNA, since highthis process to the inhibition of Db JA synthesis multiplicity passage yields such DNA molecules that band at the position of form Ic in cesium by ara-C. Two alternative mechanisms wer e considered chloride-PDI (Hiscott, Yu, and Cheevers, unto account for the synthesis of for]m Ic. It was published data). This possibility is being invesassumed that form I DNA mole,cules whose tigated. Figure 12 shows a diagrammatic summary of synthesis had already been initiate(d at the time of addition of cycloheximide as Mvell as those the results of this work, which assumes that initiated after the inhibition of prc)tein synthe- form II is a terminal intermediate of polyoma sis would finish replication normall' y (22). Thus, DNA replication (10). Upon inhibition of proform Ic could arise either by some alteration of tein synthesis, the synthesis of polyoma DNA is the closure of daughter molecules o r by removal limited by a reduced rate of initiation of new of superhelical turns from newly synthesized rounds of genome replication (22; step 1). Form form I DNA. The latter possibiility seemed I DNA molecules initiated in the absence of particularly interesting in view of t-he discovery protein synthesis finish replication normally, by Wang (19) of protein in Esct Lerichia coli, yielding closed-circular progeny DNA (22). Prowhich removes superhelical turns iin closed-cir- tein synthesis is also required to maintain the cular DNA without introducing permanent superhelix density of progeny DNA. In its strand scissions. Similar "untwistinig" activities absence (step 2) the synthesis of viral DNA have been found in extracts of m Duse (5) and shifts from the formation of normal form I human cells (13), as well as in nlucleoprotein progeny to form Ic, a monomeric closed-circular complexes containing replicating
