Structure of herpes simplex virus DNA: Topography of the molecule

Structure of herpes simplex virus DNA: Topography of the molecule

VIROLOGY 65, 494-505 (1%‘5) Structure of Herpes Simplex Virus DNA: Topography of the Molecule I. Absence IVAN HIRSCH,’ of Circularly Permuted ...

3MB Sizes 0 Downloads 26 Views

VIROLOGY

65, 494-505

(1%‘5)

Structure

of Herpes Simplex Virus DNA: Topography of the Molecule

I. Absence

IVAN HIRSCH,’

of Circularly

Permuted

Sequences

JOSEF REISCHIG,2 JAROSLAV ROIJBAL,’ VONKA’

Department of Experimental Virology, Institute Department of Biology, Medical Faculty

AND

VLADIMIR

of Sera and Vaccines, Prague, Czechoslovakia, of Charles Uniuersity, Plze& Czechoslouakia

Accepted January

and

3, 1975

Only linear structures were produced by re-annealing denatured herpes simplex type 1 virus (HSV)-DNA molecules, while T4-DNA molecules (known to be circularly permuted and therefore used as controls in the parallel tests) formed circles. These results suggested that HSV-DNA is a nonpermuted collection of sequences. Unfortunately, most of the linear structures observed after re-annealing were not full-length duplexes, thus making this test not quite satisfactory. Therefore another permutation test, the central region deletion experiment, was employed. Its results indicated that short fragments liberated from the ends of HSV-DNA by enzymatic treatment contained only a portion of the sequences present in the complete molecules. On the other hand, the presence of apparently all sequences was found in the short end-fragments of the control TCDNA. These data also orovide evidence that HSV-DNA molecules are not formed by circularly permuted collections of sequencies. INTRODUCTION

Different viruses contain either linear or circular DNA molecules. The linear molecules are either circularly permuted or nonpermuted collections of sequences. This topological aspect and its implications for constructing the genetic map and for the replication and recombination processes have been studied extensively in bacterial viruses (for review, see Thomas, 1967). Among animal viruses whose DNA replicates in the nuclei, adenoviruses and herpesviruses contain linear doublestranded DNA molecules. Recently it has been reported for several adenovirus types (Doerfler and Kleinschmidt, 1970; Younghusband and Bellett, 1971, 1972) that their DNA’s are nonpermuted collections of sequences. It was the aim of the present study to find out whether a similar situa’ Institute 2 Medical

of Sera and Vaccines, Prague. Faculty of Charles University, Plzen.

Copyright 0 1975 by Academic Press, Inc. All rights of reproduction in any form reserved.

496

tion exists for herpes simplex virus (HSV). Two permutation tests on HSV-DNA were used: the denaturation-annealing test (Thomas and MacHattie, 1964) and the central region deletion test (Thomas and Rubenstein, 1964). In the first test, the appearance of circles among the reannealed molecules implies that the respective DNA is permuted. The second test is based on competition between denatured fragments of whole-sized DNA molecules and short end (terminal) -fragments for bases of the immobilized whole-sized single chains. The presence of permuted sequences should be revealed by annealing approximately equal amounts of both short end-fragments and fragments of whole molecules. On the other hand, the preferential binding of whole-molecule fragments would signal the absence of permutation in the DNA sequences. In the present experiments T4 phage DNA, known to be circularly permuted

STRUCTURE

(Thomas and MacHattie, as a control. MATERIALS

AND

1964), was used METHODS

Viruses, their growth and purification. Three times plaque-purified HSV type 1, strain Kupka (isolated and kindly provided by Dr. R. Benda) was used. The virus was grown in a cell line derived from rabbit embryo fibroblasts, using a multiplicity of infection (m.o.i.) of 0.5 PFU/cell. The virus material for the present experiments was prepared using m.o.i. of 5 PFU/cell; the ratio of physical to infectious particles in the inoculum was 80. Conditions for preparation and purification of nucleocapsids were described in detail elsewhere (Hirsch and Vonka, 1974). Bacteriophage T4rII was obtained through the courtesy of Dr. Z. Neubauer. It was grown on Escherichia coli B in medium consisting of 1% tryptone (Oxoid) and 0.5% NaCl. The bacteria were grown with aeration to approximately lOa cells/ml, infected at m.o.i. of 1 PFU/cell, and incubated at 37” for 100 min before chloroform was added to complete the lysis. For radioactive labeling, bacteria were grown in Triscasamino acids-glucose medium (Thomas and Abelson, 1966) supplemented with tryptophan, 20 pg/ml. Radioactive labels, [32P]0,3-, 10 pCi/ml (Isocomerz, GDR), and [6-3H]uracil, 10 pCi/ml (Institute for Research, Production and Use of Radioisotopes, Prague), were added 10 min prior to infection. When using [3H]uracil, the bacteria were cultivated with 10 pg of uracil/ml and transferred to a fresh medium supplemented with 1 pg of uracil/ml just prior to the addition of the radioisotope. The phage was purified as described by Thomas and Abelson (1966) except that the suspension of T4 particles after the first cycle of differential centrifugation was treated with RNase, 50 pg/ml (Calbiothem, A grade), and DNase, 20 pg/ml (Calbiochem, B grade), for 20 min at 37”. DNA extraction. The pelleted HSV nucleocapsids or T4 particles were resuspended in neutral DNA buffer (1.0 M NaCl, 0.001 M EDTA, and 0.05 M Tris-HCl, pH 7.5), containing 0.5% sodium dodecyl sulphate (Sigma) and 2% sodium

OF HSV-DNA

497

N-lauryl sarcosinate (Koch and Light, Ltd., England), and gently rolled for 2 min at 60” with an equal volume of watersaturated phenol (distilled immediately prior to use and neutralized with so volume of 1.0 M Tris-HCl buffer, pH 8.0). After cooling, the phenol was removed by centrifuging at 4,000 g for 15 min. The phenol extraction was repeated twice at room temperature. The extraction procedure was completed by treatment with chloroform -isoamyl alcohol (2%, v/v). The aqueous phase was then gently layered onto the top of a sucrose gradient (lo-30%, w/w) made in neutral DNA buffer and was centrifuged in an SW 41 swinging-bucket rotor in a Spinco L2-65B centrifuge at 40,000 rpm and 20’ for 3.5 hr. Fractions containing the peak of activity and uv-absorbing material were pooled and dialyzed against 0.1 x SSC (SSC is 0.15 M NaCl, 0.015 M sodium citrate, pH 7). Denaturation and re-annealing of DNA in solution. Both HSV-DNA and T4-DNA were denatured in 0.1 x SSC with 50% formamide (Merck, reagent grade), pH 7.8, at 70” for 10 min. Afterward, DNA samples were rapidly cooled in an ice bath, diluted in 2 x SSC with 50% formamide to the final concentration of 3 pg of DNA/ml and annealed for 3 hr at 35” (T4-DNA) or 43” (HSV-DNA). Alternatively, DNA specimens were denatured in 0.1 x SSC without formamide at 100” for 10 min. Re-annealing then proceeded in 2 x SSC for 3 hr at 65” (T4-DNA) and 75” (HSVDNA). Electron microscopy. Specimen grids were prepared according to Inman and Schnijs (1970). The hyperphase solution (5 ~1) containing DNA molecules (0.5 pg of DNA/ml) in 0.1 x SSC, 50% formamide and 0.01% cytochrome c was spread onto a 1.2-ml drop of cooled, distilled water deposited in a shallow pit on the Teflon surface. After removing 0.1 ml of hypophase and 8min standing at room temperature, the resulting DNA-protein film was picked up on carbon-filmed copper grids, washed in ethyl alcohol, dried and shadowed with platinum at an angle of 6” and from a distance of 8 cm while rotating at a frequency of 40 rpm. Micrographs were

498

HIRSCH

taken in a Phillips EM 300 electron microscope at magnifications of 4,500 and 10,000. The magnification was calibrated by latex spheres (4,130 f 80 A, Tesla, Czechoslovakia) and according to the length of &IX 174-DNA treated in the same way as the sample investigated. Sucrose density gradient centrifugation. One-tenth-milliliter volumes of DNA solution (about 0.5 pg of DNA/ml) were carefully layered on the top of 5-ml linear sucrose gradient (5-20%, w/v) made in neutral DNA buffer. Centrifugation was performed in a swinging-bucket rotor MSE (3 x 5.5 ml) in an MSE 65 Superspeed centrifuge at 38,000 rpm and 20” for 130-180 min (as indicated in legends to figures). Fractions were collected through the bottom and assayed for radioactivity as described (Hirsch and Vonka, 1974). Molecular weights (MW) for native DNA were calculated from distances sedimented (D) according to the equation of Burgi and Hershey (1963) :

ET AL.

DNA, respectively) per ml of 0.1 M Tris-HCl, pH 8.0, 0.015 M NaCl, 0.003 M MgC12, 0.0015 M sodium citrate, and 3 x 10m5-diluted crude extract containing endonuclease I. The reaction was carried out for 7 min at 37” and stopped by chilling in an ice bath and by addition of EDTA to give a concentration of 0.006 M. To test for the possible presence of exonuclease activity in the crude extract, O.Ol-ml samples of the reaction mixtures were taken just before and after the digestion and mixed with 0.2 ml of bovine serum albumin, 1 mg/ml, and 0.2 ml of 10% trichloroacetic acid. After 30 min at O-4”, the precipitates were collected on Millipore HAWP filters, washed with 5% trichloroacetic acid, dried. and assayed for radioactivity. No drop in activity was noted, indicating the absence of exonuclease activity from the crude preparation used. MAK (methylated albumin kieselguhr) column chromatography. The method of Mandell and Hershey (1960) was used to fractionate DNA molecules on the basis of D,/D, = (MW,/MW,)“.3”. their molecular weights. The total volumes CsCl equilibrium gradient centrifugaof MAK columns were 0.9 ml, the total tion. Samples (0.1 ml) containing approxiamount of DNA was 10 pg. Elution gradimately 0.1 pg of DNA were mixed with 6.6 ents (total volume, 20 ml) were run from ml of 0.01 M Tris-HCl, pH 8.0, 0.001 M 0.53-0.70 M NaCl and from 0.55-0.8 M EDTA, and adjusted to a median density NaCl in the case of HSV-DNA and T4of 1.715 g/cm3 with CsCl in polypropylene DNA, respectively. The fractions were centrifugation tubes. The tubes were cen- dialyzed against 0.1 x SSC prior to their trifuged in a rotor MSE (8 x 14 ml) in an use in competition-annealing experiments. MSE 65 Superspeed centrifuge at 30,000 Re-annealing of DNA fragments on nirpm and 20” for 60 hr. Fractions were col- trocellulose filters. A slightly modified verlected from the bottom of the tubes, and sion of the method of Denhardt (1966) was the refractive index of representative frac- used for re-annealing of DNA fragments on filters. HSV-DNA and T4tions was measured (Kaplan, 1969). The nitrocellulose radioactivity was measured directly after DNA in 0.1 x SSC were heated for 10 min mixing each sample with 1.5 ml of water at 100” and quickly chilled in an ice bath. The DNA’s were diluted in 6 x SSC to give and 5 ml of Instagel (Packard). Digestion of DNA by endonuclease I. a final concentration of 0.01 pg/ml. TherePreparation and use of E. coli B crude after 1.2-ml aliquots were passed through extract containing endonuclease I was es- Millipore HAWP filters (1.3-cm diameter) Filters were sentially the same as reported by Rhoades at a flow rate 1 ml/min. et al. (1968). Appropriate conditions for washed with 20 ml of 6 x SSC, dried enzymatic treatment (crude extract dilu- overnight in vacua and baked for 2 hr at 80”. Filters with immobilized denatured tion and incubation time) were determined meby sedimentation of treated DNA samples DNA were treated with preincubation 0.02% each of Ficoll in neutral sucrose gradients. Reaction mix- dium containing av MW, 400,000), bovine tures contained 3 pg of DNA (1.5 x lo5 and (Pharmacia, 9 x 10’ cpm/pg for HSV-DNA and T4- serum albumin (Armour), and polyvinyl-

STRUCTUREOFHSV-DNA

499

pyrrolidone (Sigma, av MW, 360,000) in 3 Central Region Deletion Test x SSC for 8 hr at 65” (T4-DNA) or 70” (a) Characterization of HSV-DNA (HSV-DNA). The treated filters were preparation. To characterize the preparaplaced in glass vials (l&cm diameter). tion used with respect to the base composiBoth $H-labeled DNA (which had been tion, the HSV-DNA (purified as described treated with endonuclease I and fractionin Materials and Methods) was submitted ated on an MAK column) and the 32P- to equilibrium centrifugation in a cesium labeled whole-sized DNA molecules were chloride density gradient. The results are fragmented by sonication (in an MSE 100 shown in Fig. 2. It can be seen that the W sonic vibrator at O-4” and amplitude 6 p sedimentation profile was relatively sharp for 2 min) and subsequently heatand symmetrical with the peak of activity denatured. The mixtures of 0.01 rg of each at a buoyant density of 1.725 g/cm3. DNA sample in 0.4 ml of 3 x SSC were (6) Preparation of terminal fragments of added to each vial. Annealing proceeded DNA molecules. Endonuclease I of E. coli for 24 hr at 65” (T4-DNA) or 70” (HSVB was used to introduce double-chain DNA), after which the filters were removed breaks into HSV- and T4-DNA molecules. from the vials and both sides of the filters This enzyme is known to cleave both were washed with 1 x SSC. After drying, strands of duplex DNA at the same posithe filters were assayed for radioactivity as tion (Studier, 1965). Neutral sedimentadescribed above. The total radioactivity tion profiles of treated 3H-labeled and which failed to anneal to the filter was co-sedimenting control 32P-labeled HSVmeasured as trichloroacetic acid-insoluble and T4-DNA in sucrose gradients are activity. shown in Figs. 3 A and B. The amount of digested material was calculated from the RESULTS differences in distribution of relative activDenaturation-Annealing Test ities of treated and control DNA. Increase of activity in fractions that trail the main After denaturation and re-annealing, peak represents about 10% of the total both HSV-DNA and T4-DNA preparations distribution. This indicates, provided that were assayed by electron microscopy. Equal concentrations of HSV-DNA and digestion was a random event, that most of T4-DNA were used during re-annealing (3 the broken molecules received only one hit (Rhoades et al., 1968). The fragments of pg/ml). About 80% of T4-DNA molecules, less than half the length would thus have the length of which exceeded 45 pm, been derived predominantly from the two formed duplex circular structures. After ends of the DNA molecules. re-annealing of HSV-DNA, about 50 linear Both HSV-DNA- and T4-DNA-containstructures longer than 40 pm were obing reaction mixtures were then chromatoserved. No circles were found in the prepagraphed on MAK columns (Figs. 4 A and rations of HSV-DNA that had been originally denatured either in the absence of B). Several fractions eluting before the formamide at 100” or in its presence at a appearance of the main peak were collected and analyzed for molecular weight lower temperature, the number of fragmented molecules being lower in the for- by sedimentation in neutral sucrose gradients along with 32P-labeled untreated refermamide-treated DNA. Most of the HSVence DNA. The sedimentation profiles of DNA molecules had single-chain strucfractions obtained by MAK column chrotures with highly folded end-regions (“bushes”) at one or both ends of the matography are shown in Fig. 5. Using the structure, as shown in Figs. 1A and B. formula of Burgi and Hershey (1963) the end fragments obtained were 5, 31, 38, 46% Even though the absence of cyclic forms and 7, 26, 49% of the total length of suggested that HSV-DNA is a nonpermuted collection of base sequences, the HSV-DNA and T4-DNA, respectively. (c) Competition-annealing experiment. failure to obtain numerous full-length duIn competition-annealing experiments, the plexes after re-annealing made us employ sonically treated mixtures of 3H-labeled another permutation test.

500

HIRSCH

ET AL.

FIG. 1. Electron micrographs of two linear HSV-DNA molecules formed by annealing a sample that was of previously denatured (0.1 x SSC, 50% formamide, 70”, 10 min). Micrographs were taken at a magnification 10,000. Length of double-stranded region is 23.5 pm (A) and 40.0 pm (B), length of whole structure is approximately 40.7 pm (A) and 46.0 pm (B). Ends of double-stranded regions are indicated by arrows.

end-fragments and of 32P-labeled intact DNA (see Materials and Methods) were incubated with nitrocellulose filters carrying immobilized single chains. The method of data presentation is the same as that used by Thomas and Rubenstein (1964). The quantity, (Y, represents the fractional length of the central region deleted from

the 3H-labeled DNA and equals one minus twice the fractional length of end fragments (x). R is the ratio of 3zP to 3H activities of sonicated fragments that annealed divided by the same ratio for the fragments that failed to anneal. Results of competition-annealing experiments on HSV-DNA (Fig. 6) show that R increased

STRUCTURE

l-

oL

with the increase in (Y, indicating that short 3H-labeled fragments, derived from the ends of molecules, were not able to bind to filters as well as fragments derived from the whole molecules. Much smaller, if any, increase in R was observed in the case of T4-DNA (Fig. 6). When one break on the average was introduced into each HSVDNA molecule, R equals unity until x 2 0.08 but reaches the value R = 1.6 at x = 0.01 (data not shown). DISCUSSION

Two tests were employed for investigating the possible permutation of HSV-DNA sequences: the denaturation-annealing test and the central region deletion test. The failure to find circular forms by electron microscopy has provided evidence that the HSV-DNA is nonpermuted. Unfortunately, most of the linear structures observed after re-annealing were not fully duplex. Two explanations might be offered

P

10 20 30 FRACTION NUMBER

40

FIG. 2. Neutral CsCl gradient analysis of HSVDNA. Purified 3H-labeled HSV-DNA was analyzed by equilibrium sedimentation in CsCl for 80 hr in an MSE rotor (8 x 14 ml) at 30,000 rpm and 20”. Fractions were collected by puncturing the bottom of the tube, and buoyant densities were calculated from refractive indexes.

I

20

,

,

1

I

10

15

20

I

I

I

15

20

A

L 32P --*--

501

OF HSV-DNA

I

5

25 FRACTION

5

10

25

NUMBER

FIG. 3. Effect of treatment with endonuclease I on the sedimentation of HSV-DNA (A) and T4 DNA (B). 3H-labeled HSV-DNA (4.5 x lo5 cpm/3 pg/ml) and T4-DNA (2.7 x lo5 cpm/3 pg/ml) were treated for 7 min with 3 x lo-‘-diluted crude extract of E. coli B containing endonuclease I. Aliquots (10 ~1) were carefully removed, mixed with the same amount of intact ‘T-labeled HSV-DNA (A) and T4-DNA (B) diluted in 0.1 ml amounts of neutral DNA buffer and gently layered on the top of 5-20s (w/v) linear sucrose gradient in neutral DNA buffer. After 150 min (A) and 130 min (B) centrifugation in a swinging-bucket rotor MSE (3 x 5.5 ml) at 38,000 rpm and 20”, the fractions were collected and assayed for radioactivity. For a better comparison of treated and control DNA, the percentage of total activity was plotted on the ordinate.

502

HIRSCH

25

30

I

I

35

40

I

I

45

50

FRACTION

ET AL.

15

I

I

1

20

25

30

NUMBER

FIG. 4. Fractionation of endonuclease I-treated HSV-DNA (A) and T4-DNA (B) on MAK columns. Ten micrograms of 3H-labeled HSV-DNA or T4-DNA treated with endonuclease I (see Fig. 3) were loaded on an MAK column and eluted by an NaCl gradient as described in Materials and Methods. Fifty-nine fractions (0.33 ml each) from the HSV-DNA-loaded column were collected and 1 ml of 0.1 x SSC was added to each fraction. Thirty-two fractions (0.66 ml each) from the T4-DNA-loaded column were collected. Ten-microliter portions were withdrawn and assayed for radioactivity. Fractions 30-33 (A) and 19-21 (B) were dialyzed against 0.1 x SSC and selected for further experiments.

for this observation. First, single-chain polynucleotides were cleaved under the denaturation or re-annealing conditions due to the presence of heat-labile sites in HSV-DNA (Gordin et al., 1973). Second, a collection of HSV molecules could consist of molecules of varying length, as a consequence of the different DNA content in infectious and defective particles. The latter possibility seems to be less probable because of the isolation procedure used and because the HSV-DNA extracted formed a single and relatively sharp peak in sucrose and caesium chloride gradients. Although one can imagine that partially singlestranded circles might be formed even by incomplete DNA molecules if circularly permuted, the low frequency of fully duplex structures in our preparation made it difficult to consider the results as unequivocal evidence for t.he absence of circular permutation in HSV-DNA. It was mainly

for this reason that we decided to use, in addition, the central region deletion experiment. Before discussing the latter test, some other electron microscopic observations should be briefly mentioned. Many of the duplex molecules had single stranded highly folded regions, “bushes,” at one or both ends as shown in Figs. 1A and B. Recently Sheldrick and Berthelot (1974) reported the presence of terminal-redundant and internal-inverted repetitive regions in intact single-stranded HSV-DNA molecules. A large proportion of these molecules carried a bush at one end. They suggested that mispairing among reiterated sequences present within the terminal-redundant and internal-inverted repetitive regions was responsible for formation of bushes. The presence of bush structures in our material may be due to the same factors: however, different conditions of

STRUCTURE

503

OF HSV-DNA .-*--

-c

2

6

1

3

0

0

2

6

1

3 N ?

%

j0

0g

22

6$

1

3

0

0

2

6

1

3

0

10

20

10 FRACTION

20

0

NUMBER

FIG. 5. Neutral sucrose gradient analysis of HSV-DNA and T4-DNA fractions obtained by MAK column chromatography. 3H-labeled HSV-DNA and T4-DNA were treated with endonuclease I (see Fig. 3) and fractionated on MAK columns (see Fig. 4). HSV-DNA fractions 30-33 and T4-DNA fractions 19-21 were dialyzed. Aliquots containing about 1,000 cpm were carefully withdrawn and mixed with approximately the same quantity of homologous untreated SzP-labeled HSV-DNA or T4-DNA used as sedimentation markers. All samples (0.1 ml) were layered on the top of 5-ml neutral sucrose gradients (5-20’70, w/v) and centrifuged in a swinging-bucket rotor MSE (3 x 5.5 ml) at 38,000 rpm and 20” for 180 min (HSV-DNA) or for 130 min (T4-DNA). Fractions were collected through the bottom and assayed for radioactivity. The sedimentation profiles of HSV-DNA fractions 30-33 are shown in (A)-(D); the TCDNA fractions 19-21 are shown in (F)-(H). The sedimentation profiles of the mixture of untreated 3H-labeled HSV-DNA and 32P-labeled T4-DNA are shown in (E). Arrows indicate the peak of activity of the respective fragment and the position of the sedimentation marker that was used for calculating the fraction length of the fragment.

our experiments, mainly the presence of fragments of HSV-DNA in the reassociation mixture, preclude explaining our results entirely on the basis of the model proposed by Sheldrick and Berthelot (1974). We suppose that in the diluted DNA samples, as used in our experiments, the rate of intramolecular pairing is higher

than the formation of fully duplex structures from fragments of different molecules. The observation that the enzymatically liberated short end-fragments did not anneal with the immobilized single-stranded DNA equally as well as the fragments derived from the whole molecule is militat-

504 HSV GA a-+--

DNA

1 ’



HIRSCH

ET AL.

internal sequences complementary to the terminal sequences in HSV-DNA (Sheldrick and Berthelot, 1974). Their presence could influence the outcome of the reaction in two ways. First, the rate of annealing which occurred between the 3H-labeled terminal fragments and the 32P-labeled whole-sized DNA molecule fragments in the reaction mixture should increase. Second, more sequences complementary to the 3H-terminal fragments should be available on the filter-bound unlabeled DNA. Both these conditions should result in the decrease of the R values. It is therefore to be expected that the R values determined for HSV-DNA would be even higher if the internal repetitive sequences complementary to the terminal redundancy were not present in this molecule. The evidence presented in this paper strongly suggests that HSV-l-DNA is nonpermuted. Further support for this is provided by the construction of the partial denaturation map (Reischig et al., 1975 (accompanying paper)). The biological significance of this finding is unknown. It does not preclude the existence of cyclic intermediates in the replication of HSV-DNA; the recent demonstration of terminal redundancy (Sheldrick and Berthelot, 1974) makes their formation possible. It is believed that clarifying this point, in addition to its importance for understanding the topology of the genetic map and the replication cycles of HSV, may also contribute to elucidating the significance of virus genome-circularity for the integration of the genome into the host cells.

I

1

,

I

04

06

08

1.0

I8 16 t

00

02

FIG. 6. Competition between the short end-fragments and the whole-molecule fragments in annealing experiment. SH-labeled DNA’s were treated with endonuclease I (Fig. 3) and fractionated on MAK columns (Fig. 4). Molecular weight and fraction length (I) of the DNA fragments were determined by sucrose gradient centrifugation (Fig. 5). Fragments of various sizes were mixed with an equal quantity of unbroken 32P-labeled DNA. The samples (0.02 pg total DNA) were sonicated and heat denatured and then incubated with nitrocellulose filters loaded with 0.01 pg of unlabeled single chains as described in Materials and Methods. R expresses sT/sH ratio of the DNA which annealed with the filter-bound unlabeled DNA divided by the szP/aH ratio of the unbound DNA. All values were corrected for the unspecific binding of the blank filters and for background. The quantity (Y equals one minus twice the fractional length (x) of the ‘H-labeled DNA used (x I 0.5).

the permutation of sequences in HSV-DNA. It should be kept in mind, however, that this test cannot rule out the presence of a limited number of preferred permutations in the collection of the DNA molecules. Moreover, the results of the competition-annealing experiments could be influenced by the ability of the MAK columns to fractionate nucleic acids not only on the basis of their molecular weight but also according to their base composition (Sueoka and Cheng, 1962) and by the possible presence of other endonucleases in the crude extract of E. coli B used as the endonuclease I preparation. The possibility of such interference is diminished, but not ruled out, by the control tests on T4-DNA. It may be recalled that Rhoades et al. (1968) used similar experimental conditions without observing untoward effects. It may also be questioned to what extent the results of central region deletion experiments were altered by the presence of ing against

ACKNOWLEDGMENTS The authors express their indebtedness to Dr. L. Pivec for his valuable advice on MAK column chromatography and Dr. A. Krekulova for help in preparing phage lysates. They also thank to Dr. P. Sheldrick for kindly supplying them with a preprint of the paper by himself and Dr. N. Berthelot. REFERENCES BURGI, E., and HERSHEY, A. D. (1963). Sedimentation

rate as a measure of molecular weight of DNA. Biophys. J. 3, 309-321. DENHARDT, D. T. (1966). A membrane-filter technique for the detection of complementary DNA. Biochem. Biophys. Res. Commun. 23, 641-644.

STRUCTURE DOERFLER, W., and KLEINSCHMIDT, A. K. (1970). Denaturation pattern of the DNA of adenovirus type 2 as determined by electron-microscopy, J. Mol. Biol. 50, 579-593. FRENKEL, N., and ROIZMAN, B. (1972). Separation of the herpes deoxyribonucleic acid duplex into unique fragments and intact strand on sedimentation in alkaline gradients. J. Viral. 10, 565-572. GORDIN. M., OLSHEVSKY, U., ROSENKRANZ, H. S., and BECKER, Y. (1973). Studies on herpes simplex virus DNA: Denaturation properties. Virology 55, 280-284. HIRSCH, I., and VONKA, V. (1974). Ribonucleotides linked to DNA of herpes simplex virus type 1. J. Virol. 13, 1162-1168. INMAN, R. B., and SCHN~S, M. (1970). Partial denaturation of thymine and 5-bromouracil-containing X DNA in alkali. J. Mol. Biol. 49, 93-98. KAPLAN, A. S. (1969). Isopycnic banding of viral DNA in cesium chloride. In “Fundamental Techniques in Virology” (K. Habel and N. P. Salzman, eds.), pp. 487-495. Academic Press, New York. MANDELL, J., and HERSHEY, A. D. (1960). A fractionating column for analysis of nucleic acids. Anal. Biochem. 1, 66-77. REISCHIG, J., HIRSCH, I., and VONKA, V. (1975). Structure of herpes simplex virus DNA: Topography of the molecule. II. Partial denaturation map.

Virology 00, 000-000. RHOADES, M., MACHAWE, L. A., and THOMAS, C. A. (1968). The P22 bacteriophage DNA molecule. I. The mature form. J. Mol. Biol. 37, 21-40.

OF HSV-DNA

505

P., and BERTHELOT, N. (1974). Inverted repetitions in the chromosome of herpes simplex virus. Cold Spring Harbor Symp. Quant. Biol. 39, in press. SUEOKA, N., and CHENG, T. (1962). Fractionation of nucleic acids with methylated albumin column. J. Mol. Biol. 4, 161-172. STUDIER, F. W. (1965). Sedimentation studies of the size and shape of DNA. J. Mol. Biol. 11, 373-390. THOMAS, C. A. (1967). The rule of the ring. J. Cell. Physiol. 70, Suppl. 1, 13-34. THOMAS, C. A., and ABELSON, J. (1966). The isolation and characterisation of DNA from bacteriophages. h “Procedures in Nucleic Acid Research” (G. L. Cantoni and D. R. Davies, eds.), pp. 553-561. Harper and Row, New York. THOMAS, C. A., and MACHAITIE, L. A. (1964). Circular T2 DNA molecules. Proc. Nat. Acad. Sci. USA 52, 1297-1301. THOMAS, C. A., and RUBENSTEIN, I. (1964). The arrangement of nucleotide sequences in T2 and T5 Biophys. J. 4, bacteriophage DNA molecules. 93-106. YOUNGHUSBAND, H. B., and BELLETT, A. J. D. (1971). Mature form of the deoxyribonucleic acid from chick embryo lethal orphan virus. J. Viral. 8, 265-274. YOUNGHUSBAND, H. B., and BELLEIT, A. J. D. (1972). Denaturation pattern of the deoxyribonucleic acid from chicken embryo lethal orphan virus. J. Viral. 10, 855-857. SHELDRICK,