The major herpes simplex virus DNA-binding protein holds single-stranded DNA in an extended configuration.

(1)

0022-538X/83/050661-06$02.00/0

The

Major

Herpes

Simplex Virus DNA-Binding

Protein Holds

Single-Stranded

DNA

in

an

Extended

Configuration

WILLIAM T.RUYECHAN

DepartmentofBiochemistry, Uniformed Services UniversityoftheHealth Sciences, Bethesda, Maryland 20814

Received 2 December 1982/Accepted 22 February 1983

Properties of the major DNA-binding protein found in herpes simplex

virus-infected cells were investigated by using a filter binding assay and electron

microscopy. Filter binding indicated that the stoichiometry of binding of the

protein with single-stranded DNA is approximately 40 nucleotides per protein

moleculeatsaturation. Strongclustering of the protein in DNA-protein

complex-es, indicative ofcooperative binding, was seen with the electron microscope.

Measurements ofsingle-stranded fd DNA molecules saturated with protein and

spread for electron microscopy by using both the aqueous and formamide

spreadingtechniques indicated thatthe DNAis held inanextendedconfiguration

witha base spacing of -0.13nmperbase.

Protein synthesis in herpes simplex virustype

1(HSV-1)-infected cells is coordinately

regulat-ed and sequentially ordered. Three groups of

polypeptides have been designated in order of

their temporalappearance. They include the a

group,which doesnotrequire prior viral protein

synthesis for expression, the a group, which

includes proteins involved in viralDNA

metabo-lism, and thefygroup,which is madeup

primari-ly of structuralproteins found intheintact virion

(11). The 1 polypeptidesaresynthesized during

the peak period of viral DNA synthesis and include the viral DNA polymerase and the major herpesvirus DNA-binding protein, which has been designated infected cell polypeptide 8

(ICP8) basedonitselectrophoreticmobility

rela-tivetootherproteinsfound in infected cells (12).

Estimates of the molecular weight of ICP8

have ranged between 128,000 and 135,000.

DNA-cellulosechromatography ofradioactively

labeled infected-cell extracts has shown that

ICP8binds moretightly to single-stranded than

todouble-stranded DNA(4,15). Recent

experi-mentsindicate that thisproteinmaybea

group-specific antigenforherpesviruses,since antisera against ICP8 crossreact with analogous

polypep-tides found in other herpesvirus systems (20).

Finally, ICP8 has been detected in

HSV-trans-formed cells (10),andantibodytoICP8 has been

found in seraobtained from cervical carcinoma

patients (3). This report shows that this

impor-tant herpesvirus-coded protein holds

single-stranded DNAinanextendedconformation.

3H-labeled

HSV-1 DNA was the gift of

Mi-chael Bartkoski, as were initial stocks ofVero

cells and HSV-1 strain F (9). Bacteriophage fd

DNA was the gift of Lucy M. S. Chang. Calf

thymus DNA andpancreaticDNase andRNase

were purchased from Sigma Chemical Co. (St.

Louis, Mo.). Cellulosepowderand

DEAE-cellu-losepowderwerepurchased from Bio-Rad

Lab-oratories (Richmond, Calif.). Phosphocellulose

powder (Pl1) was purchased from Whatman,

Inc. (Clifton, N.J.). Preparation ofICP8wasby

theprocedure of Powelletal. (14). Atotalof15

confluent 2.5-liter roller bottles of Vero cells

were infected with HSV-1 strain F at an input

multiplicity of 20 PFU per cell. The cellswere

harvestedat18hpostinfectionandsuspended in

buffercontaining 20mMTris-hydrochloride(pH

7.5) and 0.5mM dithiothreitolata cell

concen-tration of -3 x 107/ml.Aftersonication,DNase

and RNasewereaddedtofinalconcentrations of

20 and 2 ,ug/ml, respectively. After a 30-min

incubation at 4°C, solid NaCl was added to a

concentration of 2.0 M, andtheresulting

precip-itate was removedby low-speedcentrifugation.

Purification of ICP8through chromatographyon

DEAE-cellulose and phosphocellulose was

car-ried out exactly as described by Powell et al.

(14). Each fraction from these columns was

assayed for absorbance at 280 nm andanalyzed

by sodium dodecyl sulfate-polyacrylamide gel

electrophoresis and subsequent silver staining

(19). Fractions from the phosphocellulose

col-umncontaining ICP8wereadjustedto500,ug/ml

in bovine serumalbumin, dialyzed against

low-salt buffer (50mM KCl, 20 mM

Tris-hydrochlo-ride [pH 7.5], 0.5 mM dithiothreitol, and 20%

glycerol) andapplied to a30-ml single-stranded

DNA-cellulose column (2) which had been

prewashedwithbuffercontaining bovine serum

661

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

662 NOTES

a

b

4 92

-466

445

431

FIG. 1. A 9o sodium dodecyl sulfate-polyacryl-amidegel of purified ICP8 peak fractions (a and b) from a single-stranded DNA-cellulose column. The protein was visualized by meansofa silverstaining procedure(10). Molecularweight markers,in descend-ingorder, werephosphorylase B,bovineserum

albu-min, ovalbualbu-min,andcarbonicanhydrase,

albumin. The column was eluted stepwise with

0.05, 0.1, 0.2, 0.4, and 1.0 M KCl. ICP8 was

obtained (-50,ug/109 cells) inthe0.4and1.0 M

fractions,dependingonthebatch of

DNA-cellu-lose. Theproteinwasthendialyzed against0.01

M Tris-hydrochoride (pH 7.5)-0.001 M EDTA (TEbuffer) andquantitated by using a Bio-Rad

proteinassaykit.

ICP8-fd DNA complexes were prepared for

electron microscopy by

mixing

thetwo

compo-nents atavarietyofweightratios.Sampleswere

brought to 80 ,ul with TE buffer, against which

the protein and DNA were also dialyzed. The

DNA waspresent at aconcentrationof 1 ,ug/ml

in all samples. ICP8 was present at

concentra-tions of2, 5, and 10,g/ml. The 80-,u mixtures

were incubated at4°Cfor 10minandthen fixed

with 8% glutaraldehyde (1.25

RI/80

IlI)

at room

temperatureforan additional10min. The

com-plexes were then spread for electron microscopy by either the aqueous ammonium acetate tech-nique or the formamide techtech-nique (7, 17).

Am-moniumacetate was presentin both the

hyper-phase and the hypophase at a concentration of

100 mMintheaqueoustechnique. The

complex-es wereobserved and photographed at

magnifi-cations of9,800x and 16,000x, using a Zeiss

EM 10A high-resolution transmission electron

microscope.

A solution of sonicated 3H-labeled HSV-1

DNA(80,000dpm/,ug)washeatedto100°Cfor5

minand icequenced to obtain a stocksolutionof

single-strandedDNAfor filterbindingassays.A

0.02-pg portion of the DNA was mixed with

increasing amounts of ICP8. The samples were

broughtto100 pI withTEbuffer and incubated

at 4°C for 15 min. After incubation, an 80-pA

aliquotwaswithdrawn and adsorbedtoan

0.45-,um-pore-size nitrocellulose filter witha Hoefer

multichannel filtration manifold. The filtration

speed was 4 ml/min. The filters were then

washed with 5 ml each of cold TE buffer and

90% ethanol,airdried,and counted.

Sodiumdodecyl sulfate-gel electrophoresisof

purified ICP8 and subsequent silver staining

indicated the presence ofa single polypeptide

species with a molecular weight of -128,000

(Fig. 1). Filter binding assays carried out as

^- .015-/

0 4 .010' .

z

.005-.04 .08 .12 .16 .20 .24 .28 .3 2

[image:2.490.121.167.78.248.2]

IcP8 pg

FIG. 2. Typicalresults fromafilterbinding assayindicating saturation (maximum retention) ofsonicated,

heat-denatured, 'H-labeledHSV-1 DNA with ICP8ata10:1protein-to-DNA weightratio. The HSV-1 DNA had

anaveragelengthof--400nucleotidesasdeterminedbyelectronmicroscopy(datanotshown).

J. VIROL.

http://jvi.asm.org/

[image:2.490.111.400.469.642.2]

(3)

b

'4*t{'

;*

'

v

'

05 Mm

;. >a¢#to

-0

H

-.

.

~~~~~.

;,

C.,~~~~~~V

FI. 3. Eletro

mirorpho

.ICPdcmlxs.prpae ata,\prti/N wegh rai fv: h

nonanomditrbuio ofprtin is evdne byth prsec of bot extended, ful sauae cice an

*~ v

*

itw@-FIG. 3. Electron micrograph ofICP8-fd complexes prepared at a protein/DNA weight ratio of5:1. The nonrandom distribution of protein is evidenced by the presence of both extended, fully saturated circles and

highly collapsed, irregularly shapedmolecules (arrows) with littleor no proteinbound. Thecomplexeswere

stainedwithuranylacetate(12). Glutaraldehydefixationwasrequiredtomaintain theintegrityof thecomplexes.

described above indicated that saturation of

sonicated,single-strandedHSV-1 DNAby ICP8

occurs ata protein/nucleic acid weight ratio of

10:1 (Fig. 2). This value is in good agreement

with results obtained by Powelletal.(14) which

indicate that ICP8/nucleic acid weight ratios of

10:1aresufficienttoallow thecomplete

denatur-ation ofpolydeoxyadenylic

acid-polydeoxythy-midilic acidduplexes under thermal and solution

conditions where the polynucleotides form a

stable doublehelix. The10:1 ratio indicates that

one molecule of ICP8covers approximately 40

bases. Thisstoichiometry is similartothat of the

Escherichia coli single-strand binding protein

(SSB), whichcoversapproximately 32 basesper

80,000-molecular-weighttetramer(18).

Observation of glutaraldehyde-fixed,

aque-ous-spread ICP8-fd DNA complexes at

subsa-turating levels of protein showsaclearly

nonran-dom distribution of theprotein (Fig. 3). Some of

the DNA molecules appear extended and fully

saturated with protein, whereas others are

col-lapsed irregularstructures typical of naked

sin-gle-strandedDNA under theaqueousmounting

conditions. The frequencyofopen circular and

collapsed structures at several

protein/DNA

weight ratioswasdeterminedby scoringa

mini-mum of 110 structures seen with the electron

microscopeateach ratio. Thefractionof

extend-ed open circular structures increased with

in-creasing amounts of ICP8 (Table 1). The

ob-served increase correlates quite well with the

stoichiometry established in the filter binding

assay andsuggests that theopenstructures are

fully saturated fdDNA-ICP8 complexes

where-asthecollapsedstructures arenaked fd DNA.A

similarclusteringofprotein molecules has been

[image:3.490.50.450.74.361.2]

observed withT4gene32proteinandSSB-DNA

TABLE 1. Frequency of open circularstructures at variousprotein/DNA ratios

Protein/ No.of

DNA ratio structures %Open %Closed (wt/wt) scored

0:1 110 0.0 100.0

2:1 114 18.4 81.6

5:1 130 55.0 45.0

10:1 121 94.2 5.8

VOL.46, 1983 NOTES 663

http://jvi.asm.org/

[image:3.490.255.448.601.672.2]

(4)

664 NOTES

complexes and isindicative ofcoperative

bind-ing(1, 17). Measurementsof 102 extended

ICP8-fd DNA complexes prepared at aprotein/DNA

ratio of10:1yieldedanaverage contour lengthof

0.85 ±0.11 ,um(Fig. 4). This length corresponds

to anucleotide spacing of0.132± 0.017 nm per

nucleotide basedon avalue of6,408 nucleotides

perfd molecule (5). This value is similarto the

base spacing of 0.18nmperbaseseeninSSB-fd

DNAcomplexes.

It is possible that the very close DNA base

spacingof0.132 nmmeasured for the complexes

wasdue to a combinationoftheaction ofICP8

and the aqueous mounting procedure used.

First, since the single-stranded DNA is

visual-izedas ahighly collapsedstructure,itis possible

thatportions of the putative saturated

complex-esdonothaveprotein boundtothem.

Converse-ly, it is possible that a least some ofthe

col-lapsed structures are partially saturated

complexes. Although the results presented in

Table 1 and the tightness of the distribution of

contour lengths in Fig. 4 argue against these

possibilities, they are nonetheless a source of

concern. Second, assuming that allof the open

circularstructures arefully saturated, the ionic

spreading medium used in the aqueous

tech-nique could alter the overall conformation ofthe

complex.

Toexamine thesepossibilities, ICP8-fdDNA

complexes werespread by usingtheformamide

technique, which allows single-strandedDNA to

bevisualizedas anextended smooth filament (7,

17). Procaryotic DNA-binding protein-DNA

complexes have beensuccessfully visualized as

thicker, denser filaments as comparedwith

na-kedDNA.Thus, ifpartially saturated complexes

were present in ICP8-fd DNA mixtures, the

formamidetechniquewould allowdifferentiation

ofprotein-boundandprotein-free regions. Inthe

second case, i.e., if the extended structures

werefully saturated, measurementsof their

con-tourlengths would determinewhetherthe

appar-ent geometry of the complexes was altered by

the spreading medium. fd DNAand ICP8were

mixed in80-,ul aliquots toyieldfinal

concentra-tions of 1 ,gofDNA and -7 ,ugofprotein per

ml. Incubation and fixation of the complexes

werecarriedout asdescribed above.After

fixa-tion,an additional0.1 ,gof fdDNA was added

to ensure thatsomenakedDNA waspresent. A

30-,ul portion ofthis mixture was then spread

from a 40% formamide hyperphase containing

cytochromec(0.3,g/ml)onto a15% formamide

hypophase. The grids were stained with uranyl

acetate, shadowed with gold/palladium, and

photographed at amagnification of16,000x.

Anexamination ofthegridsshowed that only

two types of structure were present:

well-ex-tended naked fd DNAcirclesand much smaller

[image:4.490.253.451.57.409.2]

.5 1.0 1.5

FIG. 4. Histogram of 102 well-extended ICP8-fd complexes prepared at aprotein/DNA ratio of 10:1. Photographs were takenat amagnificationof9,800x, using a Zeiss EM 10A high-resolution instrument. Negatives were projected onto a blackboard with a

lanternslideprojector, andcontour measurementsof

the molecules were taken with a Keuffel and Esser mapmeasuring device.

circularDNA-protein complexeswith

character-istically thicker, rougher-appearing contours

(Fig. 5). No obviously partial complexes were

seen. Contour measurements of 87 complexes

yieldedanaveragevalueof0.79 + 0.08 ,um.This

contourlengthcorrespondsto abasespacingof

0.125 + 0.012 nm per base. The naked fd in

contrasthad abasespacingof0.27nmperbase,

ingoodagreementwiththepreviously published

value of 0.28 nm per base with the formamide

technique (17).

The contour length and base spacing values

obtained by the two spreading techniques are

the same within experimental error, indicating

that (i) partial saturation of the fd DNA with

ICP8doesnot occurunder theconditionsused,

thus implying a high degree ofcooperativity in

J. VIROL.

http://jvi.asm.org/

(5)

VOL.46,1983 NOTES 66)

b

~

t~~ot7

~

~~~ *

j

a~~~~~~~~~ ~~~~~~ *..~~~~~~~~~~~~~y4' ~~~~~~~~~~~K~

~44*

4A~ ~ ~ ~ ~ '*

940~~ ~ ~ ~ ~ ~ ~ 4 VA, r

,Av.

14D~~~ ~ ~ ~ ~ ~~a '..$

P

FIG.5Anelctronicrogaphshwing a ICP8 dDNAcomple (arro) and nextndedpotein-ree f

DNAmolcuemuned y uin th frmaidetehniueandshdowdte-hicerwthgol/pllaiumNoe

"grainy"appearance Andrdcdcnorlnt ftecmlx

the binding reaction, and (ii) the tight base

spacing is imposedonthe DNAby ICP8, and the

complexesarestable underwidelyvarying

solu-tionconditions. These results imply that

single-stranded DNA in the aqueous environment of

the infected-cell nucleus can be heldin an

ex-tendedconfiguration by ICP8 witha

characteris-tic base spacing of -0.13nmperbase.

The results described above show that the

interaction of ICP8withsingle-strandedDNA is

similar to that observed with the procaryotic

single-strandbindingproteins T4gene32protein

and SSB. The highly cooperative nature ofthe

interaction seen with the electron microscope

would indicateanapparentassociationconstant

of >107 by analogy with studies on these

pro-caryoticproteins (13, 16). Both SSB andT4gene

32proteinare involved in DNAreplication and

recombination. Recently isolated

temperature-sensitive mutants with lesions in ICP8 have a

DNA-negative phenotype (6;D.Knipepersonal

communication), indicatinganessential role for

ICP8 in HSV DNA replication. Thus, in the

HSV system, ICP8 may play a role similar to

that of T4 gene 32 protein and SSB in the

bacteriophageT4 andE. colisystems.

I thank AnnaC.Weirforexcellent technical assistance and ChristaStanoyevitchfortypingthemanuscript.

This workwas supported bygrant C07114from the Uni-formed ServicesUniversityof theHealth Sciences.

LITERATURE CITED

1. Alberts,B.,andL.Frey. 1970. T4bacteriophagegene

32-astructuralproteinin thereplicationandrecombinationof

DNA. Nature(London) 227:1313-1318.

2. Alberts, B. M., and G. Herrick. 1971. DNA-cellulose chromatography. MethodsEnzymol.21:198-217. 3. Anzai, T.,G. R. Dreesman, R. J. Courtney, E. Adam,

W.E.Rawls,and M.Benyesh-Melnick.1975.Antibodyto herpessimplex virus type 2-inducednon-structural pro-teins inwomenwithcervicalcancerandcontrol groups. J. Natl. CancerInst.54:1051-1059.

4. Bayliss, G. J.,H. S.Marsden,andJ.Hay. 1975.Herpes simplex virusproteins:DNAbindingproteinsininfected cells and in the virusstructure.Virology68:124-134. 5. Beck, E.,R. Sommer, E.Auerswald,C. Kurz,B.Zink,

G. Osterburg, H. Schaller,K.Sugimoto,H.Sugiaski, T. Okamoto,and M.Takanami. 1978. Nucleotidesequence ofbacteriophage fd DNA. Nucleic Acids Res. 5:4495-4503.

6. Conley,A.J.,D. M.Knipe,P.C.Jones, and B. Roizman. 1981. Moleculargenetics ofherpes simplex virus. VII. Characterization ofatemperature-sensitive mutant pro-duced by in vitro mutagenesis and defective in DNA synthesisand accumulationof fy polypeptides. J. Virol. 37:191-206.

7. Davis,R.W.,M.Simon,and N.Davidson.1971. Electron microscopeheteroduplexmethods formapping regionsof basesequencehomologyinnucleic acids. Methods Enzy-mol.21:413-428.

8. Dreesman, G.R.,J. Burek, E. Adam, R. H. Kaufman, J. L. Melnick, K. L. Powell, and D. J. Purlfoy. 1980. Expression ofherpesvirus induced antigens in human cervicalcancer. Nature(London)285:591-593. 9. Ejercito, P. M., E. D. Kieff, and B. Roizman. 1968.

Characterizationofherpessimplexvirusstrainsdiffering intheir effectonsocialbehavior of infected cells. J. Gen. Virol. 3:357-364.

10. Flannery,V.L.,R.J.Courtney,and P. A.Schaffer.1977. Expressionofanearly,nonstructural antigen ofherpes simplex virus in cells transformed in vitro by herpes simplex virus.J.Virol.21:284-291.

11. Honess, R. W., and B. Roizman. 1974. Regulation of herpesvirusmacromolecularsynthesis.I.Cascade regula-tion ofthesynthesisof three groups of viralproteins.J. Virol. 14:8-19.

http://jvi.asm.org/

[image:5.490.57.440.71.294.2]

(6)

666 NOTES

12. Morse, L.S.,L.Pereira,B.Roizman,andP. A.Schaffer. 1978. Anatomy ofherpes simplex virus (HSV) DNA. X. Mapping of viralgenesby analysis of polypeptides and functionsspecified by HSV-1 x HSV-2 recombinants. J.

Virol. 26:389-410.

13. Newport, J. W.,N.Lonberg, S.C.Kowalczykowski, and P.H.vonHippel.1981. Interactions ofbacteriophage T4-codedgene32protein with nucleic acids.II.Specificity of

bindingtoDNA and RNA. J. Mol. Biol. 145:105-121. 14. Powell, K. L., E. Littler, and D. J. M. Purifoy. 1981.

Nonstructuralproteins of herpes simplex virus.II.Major

virus-specific DNA-binding protein.J. Virol. 39:894-902. 15. Purifoy,D.J.M.,andK.L. Powell.1976.DNA-binding proteins induced by herpes simplexvirus type 2 inHEp-2 cells. J. Virol. 19:717-731.

16. Schneider, R. J.,and J. G. Wetmur. 1982. Kinetics of transferof Escherichiacoli single strand deoxyribonucleic

acidbinding protein between single-stranded deoxyribo-nucleic acid molecules. Biochemistry 21:608-615. 17. Sigal, N.,H.Delius,T.Kornberg,M.L.Gefter,andB. M.

Alberts. 1972. ADNA-unwinding protein isolated from Escherichia coli: its interaction with DNA and DNA polymerases. Proc. Natl. Acad. Sci. U.S.A. 69:3537-3541.

18. Weiner, J. H., L. L. Bertsch,and A.Kornberg. 1975. The deoxyribonucleic acid unwinding protein of Escherichia coli: properties and functions in replication. J. Biol. Chem. 250:1972-1980.

19. Wray, W., T. Boulikas, V. P. Wray, and R. Hancock. 1981. Silver staining of proteins in polyacrylamidegels. Anal.Biochem. 118:197-203.

20. Yeo, J., R. A.Klllington,D. H.Watson, and K. L. Powell. 1981. Studiesoncross-reactive antigens inthe

herpesvi-ruses.Virology 108:256-266.

J. VIROL.

http://jvi.asm.org/