The herpes simplex virus Us11 open reading frame encodes a sequence-specific RNA-binding protein.

JOURNAL OF VIROLOGY,

JUlY

1990, p. 3463-3470 0022-538X/90/073463-08$02.00/0

Copyright© 1990,American Society for Microbiology

The Herpes Simplex

Virus

Us11

Open Reading

Frame

Encodes

a

Sequence-Specific

RNA-Binding

Protein

RICHARD J. ROLLERANDBERNARD ROIZMAN*

TheMarjorieB. Kovler Viral OncologyLaboratories, The University ofChicago, Chicago, Illinois 60637

Received 21 February 1990/Accepted12April 1990

Herpes simplex virus 1- and 2 (HSV-1 and HSV-2)-infected cell extracts but not uninfected cell extracts

containanRNA-bindingactivity foraninvitro-transcribedsequencefromthedomains of theHSV-1Us11and

a47genes.The transcript of thissequencehasnotbeen detected ininfectedcells. The binding issequenceand

secondary structure specific and protects approximately 95 nucleotides from RNase digestion. Analyses of HSV-1 x HSV-2 recombinants and HSV-1 deletionmutantsmapped the functionnecessaryfor activitytothe

Us11

ora47 open readingframe. The a47genewas excluded, since the RNA-binding activity is alate (Y2)

function dependent on viral DNA synthesis for its expression. The

Usll

function is theonly viral function required, since translation in rabbit reticulocyte lysateofaninvitro-synthesizedUs11 mRNAresulted in the

appearance of the RNA-bindingactivity. The product of theUs11open readingframe is associated with the

RNAprobe-proteincomplexinasmuchasinsertionofa sequenceencodingin frame 15 additional aminoacids attheC terminusof the

Usll

protein causedacorrespondingdecreasein theelectrophoreticmobility of the

bindingcomplex.

Inthispaper wereportthataprotein encoded byan open

readingframe in thesmallcomponentof theherpessimplex virus 1 (HSV-1)genome participates in the formation ofa

sequence-specific RNA-protein complex in vitro. This is the firstreportofanHSV protein withsequence-specific affinity

forRNA. Relevanttothisreport arethefollowing.

(i)The 74 known openreading frames ofHSV-1 (1, 3, 5,

24,25)formseveralgroupswhoseexpression iscoordinately

regulated and sequentially ordered inacascade fashion (10,

11). Among the products of theagenes,the firstgrouptobe expressed, isthe product of the a4gene, designatedas the

infected cellproteinno.4(ICP4). This protein bindstothe 5' untranscribed and transcribed noncoding domains of ,B and 'Yi and _Y2 genes

expressed

later in infection

(14).

Because ICP4 binds to transcribed noncoding domains of Y2genes,

thatis,genes stringently dependentonviral DNAsynthesis

for theirexpression,thereexisted theformalpossibility that ICP4bindstoRNAaswellasDNA,apossibility that isnot too farfetched in light ofthe discovery of a host protein which binds single- and double-strandedRNA and DNA in sequence-specific fashion(36). Byafortuitous coincidence,

we chose to test the transcripts of the _Y2

US11

(13) open

reading frame. While we were unable to show that ICP4 bindstoRNA, wediddiscoverthat

Usll

proteinbinds toa

transcript complementarytoitsdomain.

(ii)Theprotein productof the

Usll

genewasfirst identi-fied, using hybrid arrest oftranslation, by Rixon and Mc-Geoch (33) and was shown to be a protein with an Mr of 21,000. Subsequently, usinganantiserumdirectedagainst a

peptide sequence predicted by the nucleotide sequence of the

Usll

openreading frame,Johnson etal.(13)showedthat the

Usll

protein actuallyconsists ofadoubletwithanMrof 21,000to22,000and that theexpressionof

Usll

isregulated

as aY2ortruelategene.Johnson et al.(13)alsoreportedthat

the size ofthe

Usll

proteinvariedbetweentwoof the strains tested-an observation which had been predicted on the basis of the differencesin thesize ofthe openreadingframe found in two different strains of HSV-1 (33, 43). These

*Correspondingauthor.

results were consistent with those of Lonsdale et al. (20), whohad earlier shown thatadoublet of proteins withanMr

ofapproximately 20,000 fromanumber of HSV-1 laboratory

strains and isolates passaged a limited number of times

outside the human hostvariedsignificantlyinits electropho-retic mobilityindenaturingpolyacrylamide gels.

(iii)Ourlaboratory and otherlaboratories have reporteda

large numberof HSV-1genesthataredispensable for growth

in cells in culture. Relevant to this report are studies

demonstrating that, with one exception, all of the genes

mappingin the uniquesequences of the S component

(Us),

including the

Usll

openreadingframe,aredispensable (19).

None of the studies done to date in our laboratory or

reported elsewhere have shown that the

Usll-

HSV-1 strains are deficientfor replication in any of the cell lines

tested.

MATERIALS AND METHODS

Cells and viruses. HSV-1 strain F [HSV-1(F)], HSV-2 strain G [HSV-2(G)], and deletion mutants R325, R7023, R7039, R3630, and R3631 weredescribed elsewhere (7, 18, 23, 27, 29, 35). HeLacells (American TypeCulture Collec-tion)werepropagated asdescribed elsewhere (14).

Construction andcloningofplasmids. Standardprocedures

wereused in allconstructions described in this report (22).

The structures of plasmids pRB3881, pRB3910, and pRB4027anddetails of their constructionaregiveninFig. 1. Construction of recombinant virus. The recombinant

vi-ruses used in this study (R325, R7023, R7039, R3630, and R3631) and the one recombinant virus constructed for this studywere constructed by double homologous recombina-tion betweenthe intactDNA ofaparentvirus andplasmid DNA in which nonhomologous sequences and a selectable markerweresandwichedbetweenflankingsequenceswhose homologswerepresentin the sameorder andorientationin

the genomic DNA. The selectable marker was the HSV-1 thymidinekinase(tk)geneasdescribedbyPostand Roizman (29), except that the natural promoter of the tk gene was

replaced bythea27promoter(19). RecombinantR3630was

constructed by replacementof thecx47andUs11geneswith

3463

Vol. 64,No. 7

on November 10, 2019 by guest

http://jvi.asm.org/

3464 ROLLER AND ROIZMAN

A[

2f~

ji

B "Yin -M Deletion

-R7023

1 R7039 _

R325 U

R3631 U

C

DI

gl usl1o

F-*

(H-S2 G gD> US9 USi1

2 a4 a22 US3 US5 gE a47 x4

a. c' U c a

- r~zz~~

USi 1

a47

1111 Oris RNA

Sal SacSac Bam Rsa,Bst Rsa Sac Eco

pRB421

T7

2 SmaBst Sty Sma ThaPst

U- 1 ---, pRB3881

: ' ' ~~~~~~~SP6

3 Eco BamSac Xho Xho

pRB3910

iSal Ava Bar Sac Ect

pRB421 5bp Synthetic Tag , 07bpAva/Sac Fragment 2m/AvaCompatibleEnds|- ; va/SacCompatibleEnds

Sal Bam,Sac Eco

4, Bam/Sac 1-cutpRB421

3 US1O(truncated)

< USll(tagged)

Sal ,p,5jac§ Eco

pRB4027

FIG. 1. Sequence arrangement of the HSV-1 genome and S componentand constructionofplasmids.(A)Schematic diagram of the HSV-1 genome in the prototype arrangement, showing the unique sequences (lines) flanked by inverted repeats (boxes). (B) Expansion of the S component of the genome. Bi, Extent of deletions intheshortcomponent of deletion mutants usedinthese studies. Inthe genome of R7023, theScomponent is frozenin an invertedorientation. Thedeletion extends beyond the L-Sjunction intotheinvertedrepeatsflankingtheScomponent.B2,Genetic map of the short component for comparisonwithBi. (C)Expansionof theregion oftheS componentcloned aspRB421andconstructionof plasmidspRB3881and pRB3910.Cl, Restrictionandgeneticmaps ofpRB421 (23). Verticalbarsindicatetranscription initiation sites; thin lines indicated transcribed noncoding domains; solid boxes indicate open reading frames ot47,

Us11,

and

Us10.

a47 and Oris RNAtranscripts initiate within the invertedrepeat sequence; Us11 andUs10transcripts initiate withinUs. Restriction enzyme abbre-viations: Sac, SacII; Bam, BamHI; Bst,BstEII; Rsa, RsaI, Sal, Sall;Eco,EcoRI. Thelisting of RsaI sitesis notcomplete; only the relevant sites are indicated. C2, Construction of pRB3881 probe template. The202-ntRsa-Rsafragment containing 158 nt of

Us11

5'-untranslated sequence and 44 nt of5'-flanking nontranscribed sequence was cloned into the HincIl site of pUC8 and then subclonedas aHindIII-EcoRIfragmentintopGEM-3Z. Thin lines indicate sequences transcribed by T7 RNA polymerase; the open box indicates the minimumregionessential for binding (see text). Forprobe transcription, plasmidwaslinearized at thePstIsite and transcribed with T7 RNA polymerase toproduce a245-nt probe. Restrictionsitesused togeneratedeleted probes: Bst,BstEII;Sma, SmaI; Sty, StyI; Tha, ThaI; Pst, PstI. 5' deletionswere made by collapsing the template between thepolylinkerEcoRI site and the appropriate site intheinsert.3'deletions were made bylinearizing pRB3881 at sites within the insert. C3, Construction ofpRB3910 plasmid template for

Us11

mRNA. The650-base-pairBstEII-SacII

thecx27-tkgene inahomologousrecombination between the genome ofa tk-negative virus anda plasmidcontaining the

a27-tk gene sandwiched between the appropriate homolo-gous sequences. The recombinant was selected by

plating

the progeny of the cotransfection in medium containing methotrexate, which selects for the tk+ progeny. Recombi-nant R4027 was constructed by cotransfection of intact R3630 DNA andpRB4027 plasmid DNA containinganinsert in the

Usll

open readingframeas described in Results. In this instance, the tk gene was replaced by the HSV-1 sequences contained inpRB4027.

In vitro transcription. Plasmid templates were prepared

and transcribed with SP6 and T7 RNA polymerases as recommended by Promega Biotec, except that all RNAs were capped during synthesis with the cap analog GppG (New England BioLabs) as described previously (28). La-beled binding probe was synthesized with 10 1Ci of [32P]CTP ina 10-,ul reaction mixture supplemented with 50 ,uM unlabeled CTP. Reaction products were separated on

denaturing4% acrylamideureagelsand visualizedby auto-radiography of the wet gels. The band corresponding to full-length probe was excised, and RNA was eluted by soaking overnight in 0.5 M ammonium acetate-1 mM EDTA-0.2% sodiumdodecyl sulfate. Eluted RNA was pre-cipitated with ethanol in the presence of 10 pLg of carrier glycogen and suspended inwater.

Capped synthetic

Usll

mRNA for in vitro translationwas transcribed from EcoRI-linearized pRB3910 by the same procedure, except that the concentration of all NTPs was 0.5 mM andonly 0.5 ,uCi of [32P]CTPwasused asatracer. The product was quantitated by DE-81 filter assay as described elsewhere(22).

Assays for RNA-binding activity. For gel shift assays, labeledprobe (10,000to20,000 cpm) was treated with 5 plg of nuclear protein extract in 7

[l

ofbinding buffer (200 mM NaCl, 20 mM Tris [pH 7.9 at 4°C], 2 mM MgCl2, 0.5 mM dithiothreitol, 0.2 mM EDTA, 0.2 mMphenylmethylsulfonyl

fluoride, 20%[vol/vol] glycerol) for 20min at room temper-ature, followed by addition of heparin to 5 mg/ml and reaction for a further 10 min. Onemicroliterofloadingdye (100%glycerol, 0.4%bromphenol blue, 0.4% xylene cyanole FF)was then added, and samples were applied to a nonde-naturing 4%acrylamide gelpoured and run in 0.5x TBE(45 mM Tris, 45 mM boric acid, 1.25 mM EDTA). Protection

fragment of pRB421 containing theUs11 openreading frame, 76nt of5'-untranslated sequence, and 90 nt of3'-untranslatedsequence wasexcised, converted to blunt ends by treatment with the Klenow fragment and dNTPs, and cloned into the SmaI site ofpGEM-3Z. Linearization with EcoRIrestrictionenzymefollowedby transcrip-tion withSP6RNApolymeraseproducedUs11mRNA. Restriction site abbreviations: Eco, EcoRI; Bam, BamHI; Xho, XhoI. (D) Construction ofpRB4027 encodingtagged Us11 protein. Two in-serts were prepared asfollows. (i) The 107-base-pairAvaIl (Ava)-Saclfragment of pRB421 (D1) encodingUs11codons111to 146 was excised and purified from a low-melt agarose gel. (ii) Synthetic oligonucleotides (94 and 95 nt) wereannealedtogeneratean insert encodingUs11 codons 147 to 161followedby asequenceencoding amino acids 71 to 85 of HSV-1 ICP4 and a UAA stop codon. The annealed double-stranded insert had overhanging BamHI- and AvaII-compatible ends. These inserts were coligated (D2) into SacI/BamHI-cutpRB421 togenerate pRB4027 (D3). The newUs11 openreadingframe encoded acompleteUs11plus tag. The insertion of the tag introduced a stop codoninto theUs10open reading frame, truncatingit such that only 110 codons should be translated. The construction also resulted in the duplication of the sequences between the BamHI and Avallsites.

r---

---U.Y6

---F e

I

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.612.66.298.77.415.2]

HSV RNA-BINDING ACTIVITY 3465 assays were identical, except that before addition of

heparin,

reaction mixtures were digested with 5 U of RNase T1 for 10 min.

Preparation of cellular extracts. Small-scale nuclear ex-tracts (used for experiments shown in Fig. 5) were prepared asdescribed by Lee et al. (16) from HeLa cells infected with 10PFU per cell and harvested 20 h after infection. Large-scale nuclear extracts (used for all other experiments) were prepared as previously described (14).

Sephacryl S300 chromatography. One milliliter of nuclear extract was fractionated on a 55-ml column of Sephacryl

S300 run in binding buffer at a linear flow rate of 8

cm/h;

0.5-mlfractions were obtained.

Preparation ofcompetitor nucleic acids. Competitor RNAs were synthesized and purified in the same way as probe RNA, except that reaction mixtures contained all NTPs at 0.5 mMwithout label, and the product was excised from gels

byusingcomigrating labeled product as a standard. Purified

competitor RNAs were quantitated by

A260

Competitor DNAs were purified from low-melt agarose gels as described elsewhere (22) and quantitated by coelectrophoresis with DNAsof known quantity.

In vitrotranslation. Synthetic

Us11

mRNA (0.3 ,g) was translated in a 20-,ul reaction mixture by using a kit obtained

fromPromega Biotecaccording to the protocol supplied with the kit, except that after the suggested incubation period, unlabeledmethionine was added to 100 puMand incubation

wascontinuedfor afurther 30min. After the reaction, the

mixtureswerediluted to 50

,lA

with binding buffer and passed over a 1-ml Sephadex G-50 spin column which had been

preequilibrated inbinding buffer.

RESULTS

Infected-cell extracts contain an RNA-binding activity.

pRB3881 (Fig. 1C) contains 5'-transcribed noncoding se-quences of the Y2

US"1

open reading frame. Uniformly

labeled in vitro transcripts of each strand of the HSV-1 sequences inpRB3881 werereacted with apartiallypurified preparation of ICP4 derived from Bio-Gel A.5m

chroma-tography ofcrude HSV-1-infected cell nuclear extract (P. McAtee and B. Roizman, unpublished results), and the

reaction mixtures were electrophoretically separated on a

nondenaturing polyacrylamide gel. Only one ofthe RNA

probes,i.e.,thatcorrespondingto theoppositestrand of the

Usll

open reading frame 5'-untranslated sequence, exhib-ited a slower

electrophoretic

mobility

(Fig.

2A,lanes 1 and 2).Digestion ofthe mixture withproteinase Keitherbefore or afterthe reaction

(Fig. 2A,

lanes 3 and 4,

respectively)

causedthe RNA

probe

to movewith normalmobilityonthe

gel,

showing

thattheretardation ofthe

probe

wasduetothe

formation ofanRNA-protein

complex.

Because thedegreeof retardationappearedtobetoosmall to be caused by a

protein

as

large

as ICP4 and because

addition of anti-ICP4

antibody

to the incubation failed to causeafurther decrease in themobilityof theprobe,ashas been shown for ICP4-DNA complexes (14), it seemed un-likelythat the retardationof the

probe

was dueto ICP4.

RNA-bindingactivity issecondary structure

specific.

RNA

probe transcribed from

pRB3881

was

always

purified

as a single tight band from

denaturing

polyacrylamide

gels.

In

nondenaturing gels, however, it

migrated

as two

species.

The

faster-migrating

species

composed

70% ofthe label and

formed a

relatively

diffuse

band,

whereas the

slower-mi-grating

form

composed

30% of the total and formed a

relatively tight band

(Fig. 2A,

lane

1).

Elution of either

A B-

-ProltK~~

E Co E.2)

A.5m

E S S 3 E , ¢~~E c

Protein > >

I.Q cL <I )

;K3: <

_

_

_ ^

_ _

.w.. _

..:..._

I.

-u

42

h

123 4 1 2 3 1 2 3 4 5

FIG. 2. Demonstration of bindingto specific probe and protec-tionof probe fromRNaseT1 digestion.(A)Autoradiographicimage ofpRB3881 T7 transcript probe reacted as describedin Materials and Methods with binding buffer (lane 1), 7

[lI

ofBio-Gel A.5m fraction(lane 2),7,ulofA.5mfractionpreviouslydigested with 200

p.g

ofproteinaseK(Prot. K)permlfor30minat25°C (lane 3),or7 ,ulof Bio-GelA.5mfraction subsequentlydigested withproteinase K (lane 4). (B)Autoradiographic image of pRB3881 T7transcript probe reactedwith 7 ,ul of Bio-GelA.5mfraction(lane 1), 5 ,ulof HSV-1(F)-infectedcell nuclearextractfollowedbyRNase

T,

diges-tion (lane 2), or 5 jg of HSV-2(G)-infected cell nuclear extract followedbyRNase T1 digestion (lane 3). Arrowheads indicate the positions ofspecific RNase-resistant RNA-protein complexeswhich wereexcised from the gel.(C)Autoradiographic image ofdenaturing electrophoresis of RNAs isolated from the complexes shown in panelB. Lane 1,pRB3881 T7transcript probe RNA; lane 2,RNA fromA.5mfractionbinding complex;lane3,protected RNA frag-ment from HSV-1(F) binding complex; lane 4, protected RNA fragment from HSV-2(G) complex; lane 5, MspI-digested, end-labeled pGEM-3Z molecular weight (MW) standards (fragment lengths shownatright).

species from gel slices followed

by

electrophoresis

on an-other

nondenaturing gel

resulted in thereappearanceofboth

species in the same ratio (not

shown).

The

probe

species

were interconvertible and

apparently

represent different secondary structuresadopted by the RNA

probe.

Only

the slower

species

formed

complexes

with the infected cell proteins (Fig. 2A, lanes 1 and 2;

Fig.

3B, compare lanes 1 and 2through 14).

RNA-binding

activity

issequence

specific.

The RNAs tested todate include

transcripts

of both ends of the

unique

short

region ofthe HSV-1 genome and sequences from the

5'-untranslated

region

ofthe HSV-1 late gene

encoding

the a gene

trans-inducing

factor

(aTIF).

In

all,

severalkilobasesof RNA sequences have been

tested,

and no other sequence

VOL. 64, 1990

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.612.353.533.73.382.2]

3466 ROLLER AND ROIZMAN

B

CompetitorRNA Competitor DNA

Pst 3881 t,nci {,Ta3881 Vc.cI.r i.

-CO t:30G

qI'O1- _%£tCMCn

.

~~~~CM

M v iC13 -> w _{C) CO L. n}

[image:4.612.129.460.77.247.2]

1t1112 13 4 15 1

FIG. 3. Autoradiographic images of complexes formed betweenA.5m fraction protein andpRB3881 T7transcript probe RNA in the

presenceofvarious competitor RNAs andDNAs. Arrowheads indicate thepositionsof theRNA-protein complexes. (A) Competition with

specific (lanes 3 to7) and nonspecific (lanes8 to17)RNAtranscripts. Lane 1, Probe without protein;lane2, probeand proteinwith no

competitor. (B) Competition with specific (lanes 3to8)and nonspecific (lanes9to 14)double-stranded DNAcompetitors. Lane 1, Probe without protein; lane 2, probe and protein without competitor.

has been found to bind. Transcripts of various truncated versions of the HSV-1 sequences in pRB3881 probe have

also been tested forbinding. 5' deletions upto theStyI site boundaswellasthecomplete probe,but additional deletions to the3'-mostSmaI site abolishedthecapacity of the RNA tobind. We have been unabletodeleteanysequencesfrom the 3' end and retain binding ability; linearization of pRB3881 with ThaI yielded a nonbinding transcript shorter by only7nucleotides (nt).The distance between the 3'-most SmaI site and the ThaI site is 87nt,showingthat the smallest fragment capable of binding isatleast this size.

Sequence specificity oftheRNA-binding activitywasalso

demonstrated in competition assays (Fig. 3A). Specific

probe was mixed with various amounts of unlabeled

com-petitor RNAs, either specific (Fig. 3A, lanes 3 to 7) or

nonspecific (Fig. 3A, lanes 8 to 17). The nonspecific

com-petitors used were (i) the T7 transcript of ThaI-linearized pRB3881 and (ii) a 269-nt T7 transcript of the promoter-containing PvuII fragment of pGEM-3Z which also was

nonbindinginthe mobility shiftassay.The salient features of

the resultwereasfollows. First, the specific competitorwas

mosteffective, but allof the competitors decreased binding to some extent, and indeed it appears that the specific

competitor isonly three- tofivefold better than the others. The smalldifferencemaybe deceptive in that onlyafraction

of the 3881 competitoriscompetitive because of secondary structure. The two nonspecific competitors, on the other

hand, formed single, discrete bands in nondenaturing gels and may be completely competitive. If this is taken into

account, the specific competitor was 10- to 20-fold more

effective. Second,alargemassexcessofspecific competitor

(about 300-fold) was required before competition was

ob-served. The reasonfor this is unclear, since labeled probe

was apparently inexcess overbinding activity (i.e., only a

little morethan half of the labeled probe was bound inthe

absence ofcompetitor).

Competitionswere also done with both specific and

non-specific double-stranded DNA to determine whether the activity is RNA specific. Neither the 240-base-pair EcoRI-PstIinsertfragment of pRB3881 (specificcompetitor)northe

379-nt PvuIIfragment of pGEM-3Z (nonspecificcompetitor) competed effectively for the RNA-binding activity (Fig.3B).

RNA-binding activity is present only in infected cells. Nu-clearextractsfrom infected and uninfectedcells were chro-matographedon aSephacryl S300gelfiltrationcolumn,and column fractionswere assayedforRNA-binding activity by gel mobility shift(Fig. 4). RNA-binding activitywaspresent infractions of infected cellextract (Fig. 4A), peaking near

the void volume and tailing well into the column. This activitywas absent in all fractionsof uninfected cellextract (Fig. 4B), indicating that some function necessary for this activityis either induced orencodedby HSV-1.

AnalysesofRNA-binding activityincrude nuclear extracts. Thegel shiftassay described above didnotpermitassayof RNA-binding activity in crude nuclear extract, since

com-peting nonspecific binding (e.g., by heterogeneous nuclear ribonucleoproteins)masked the shift. We found that itwas

possible toassaytheRNA-binding activity in crude extract by a protection assay as described by Leibold and Munro

(17),inwhich thebindingreactionwasfollowedby digestion

with RNase T1, which destroyed nonspecific complexes. Mixtures ofprobeand either A.5mfraction (Fig. SA,lane4)

orcrude HSV-1-infected cell nuclear extract(Fig. SA,lane 5) resulted in the formation of two complexes between proteinandprotectedRNAwhichwerenotpresenteitherin the absence ofprotein (Fig. SA,lane2)orinthepresenceof

mock-infected cellprotein (Fig. SB, lane 1), confirming that the binding and protecting activity is either induced or

encodedbyHSV. Theamount of theuppercomplex varied

from one experiment to the next, and it was occasionally

absent(e.g., Fig. SB,lane3). Bandsseen evenintheabsence

ofprotein presumably resulted from self-protection of the probe due to secondary structure. HSV-2(G)-infected cell extracts also reactedwith the RNA probe and yielded two specific complexes (aswithHSV-1, only the faster-migrating complex was invariably present) which migrated slightly moreslowlythan their HSV-1counterparts(Fig.SB,lane 2). RNA-binding activity requiresthe product of theU1llopen

readingframe. Inpreliminary experiments,HSV-1 x HSV-2 intertypic recombinants were used to map the viral gene

responsible for the mobility difference between the HSV-1 and HSV-2 complexes to the unique sequences of the S

component. We thentested four deletion mutations ingenes

mappingin the S component (Fig. 1B). Nuclearextracts of

A

J. VIROL.

i

-0

31- 'L, A&MAAn

W**oioiow*WF,Iwwio:w

K

on November 10, 2019 by guest

http://jvi.asm.org/

HSV RNA-BINDING ACTIVITY 3467

A A HSV-1

SephacrylS300 Fraction

to NDC-e

**

O-N.

-%t*_w*F**X*****sX

fn v r, co m

,

Oj

~~~~~~~~.).

-n,rXa,SNC .co v ' coQKN

B

00 LL

E E.0 COC'

( LE E >

eC>

°'

N>_ED

tL

B Mock

SephacrylS300 Fraction

;;'4C,Lt CS rS e , =

[image:5.612.64.297.76.533.2]

<lCvvesrccrvvv- n_ ;,,

FIG. 4. Autoradiographic images of complexes formed between

pRB3881 T7 transcript probe and Sephacryl S300 fractions of HSV-1(F)-infected cell nuclear extract (A) or mock-infected cell

nuclearextract(B). The position of RNA-protein complex migration

wasdetermined by running acomplexformed between the probe

and the A.5mfraction(markedwitharrowheads)inpanelsAandB

(lanes labeled A.5m). PanelB, lane S300,showscomplexes formed

between probe and protein pooled from fractions 39to 49 of the

fractionationdepictedinpanelA.

two of these, mutants R7023 (Fig. 5A, lane 7) and R3630 (Fig. SB,lane4),failedtobindthe RNAprobe.Thecommon

property of both mutants is the absence of(x47 and Us11

openreadingframes.

To determine which of these two open readingframes is required,wetookadvantageof the observationthat

Usll,

a

Y2gene, requiresfor itsexpressionviralDNAsynthesisand

is not expressed in the presence ofphosphonoacetate, an

inhibitor ofDNAsynthesis, whereas x47 isexpressedin the

presenceof this inhibitor. Sinceextractsof cells infected and

C z

z-+ E:Uu iL 'C E

o c o+ 1:_

I) o: ti U)(

Z j I

^

i!

42 -1

D

t--

r(

> C-)

-n

*Wu .

1 2 3 4 5 6 7 8 1 2 3 4 5

FIG. 5. Autoradiographic images of complexes formed between pRB3881 T7 transcript probe and various protein samples in gel shift (panelA, lanes 1 and 3) and RNaseT, protectionassays (all other lanes).Thepresence of protein or state of infection of the cells from whichextractwas prepared is indicated above each lane. The genes deletedinviralmutants R325,R7023, R7039, and R3631 are shown in Fig. 1B. Viral mutant R3630carries the same deletion as R3631 but is tk negative. (A) Positions of the HSV-1-specific protected complexesareindicatedbyarrowheads. Note that although all lanes arefromthe same autoradiograph, only lanes 1 to 4 were adjacent in the original. (B)Positionsof the HSV-1- and HSV-2-specific com-plexes areindicated byarrowheadsmarked 1 and 2, respectively. The lanes were not adjacent in the original autoradiograph. PAA, Phosphonoacetic acid. (C)Lane 1, no protein; lane 2, protein from rabbitreticulocyte lysatein which noexogenousmRNA was trans-lated; lane 3, protein from rabbit reticulocyte lysate in which pRB3910 SP6-transcribed Us11 mRNA was translated; lane 4, protein fromHSV-1(F)-infectedcell nuclear extract. Thepositionof thebinding complex is indicatedwith an arrowhead. Thelaneswere not adjacent in the original autoradiograph. (D) Comparison of complexes formed in extracts of HSV-1(F)- and R4027-infected cells. Positions oftheHSV-1(F)-andR4027-specificcomplexesare indicated byarrowheads.

maintained in the presence of phosphonoacetate didnotbind the RNA probe (Fig. 5B, lane 5), we conclude that the

RNA-binding activity requires

Usll.

The

Usll

open reading frame product is the only viral protein required for RNA-binding activity. Invitro

transcrip-tion-translation was used to demonstrate that the

Usll

functionis theonlyviral functionrequiredfor the formation of the RNA-protein complex. In vitro transcription of EcoRI-linearized pRB3910 (Fig. 1C) with SP6 polymerase produced a710-nt mRNA. When translated in rabbit reticu-locyte lysate, thismRNAproducedaproteinwhichmigrated

in denaturing polyacrylamide gel electrophoresis as a dou-blet with an Mr of approximately 22,000 to

23,000

(not shown), as expected for

Usll

protein (13).

In an initial

experiment,incubationof

binding

probewith

lysate

in which

Usll

had been translated didnot result in the formation of anyspecific complex.Thisprovedtobe dueto an

inhibitory

factor in the translation

mix,

since

mixing

the crude

trans-VOL.64, 1990

*4 .4

$-.4.;'Al

;?, -T .1 I..

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.612.326.555.76.322.2]

3468 ROLLER AND ROIZMAN

lation product with HSV-1-infected cell extract also pre-vented complexformation. Passage of the translation mix-ture over a Sephadex G-50 spin column in binding buffer removed theinhibitory factor and resulted in complex for-mation with the RNA probe (Fig. SC, lane 3). Complexes

were notformed with protein fromlysate in which

nothing

(Fig. SC, lane 2) or an irrelevant mRNA (either Brome mosaic virus RNA or synthetic HSV-1 aTIF mRNA; not shown) had beentranslated.

The

Usll

open reading frame product participates in the RNA-binding complex. To determinewhether the

Us11

pro-teinparticipates in thebinding complex,weinserted into the 3'terminus ofthecoding domainof

Usll

asequencecoding

a15-amino-acid peptide in framewith the

coding

sequenceof

thegene, asillustratedinFig. 1D. Thechimericgene,cloned aspRB4027, was recombined into the genomeof

recombi-nant

R3630

in which the a-47 andportions ofthe

Us11

open

reading frame were replacedby thethymidine kinasegene.

Previous studies have shown that the 15-amino-acid se-quencecontainsanepitope foramonoclonalantibodyH1091 to ICP4 (12).

Therecombinant virus (R4027)selectedfor thethymidine kinaseminus phenotypeexpressed aproteinwhichmigrated

more slowly in denaturing polyacrylamide gels than the authentic protein, and, moreover, after transfer to a

nitro-cellulose sheet, the protein reacted with the monoclonal

antibody H1091 (not shown). Extracts ofcellsinfected with R4027bound to the RNA(Fig. SD, lane2), and, moreover, the mobilities of the T1-resistant complexes were slightly

less than those of complexes fromHSV-1(F)-infected cells, consistent with the slightly larger size of the tagged

Usll

protein. This strongly suggests that the RNA-protein

com-plex contains

Usll

protein.

RNA-binding activity protects a large piece of the probe from RNase digestion. Thedeletionstudies described above

suggestedthat the minimum RNAprobesize wasin excess of 87nt. Itseemedunlikelythateach of these 87or more nt isrequired for recognitionby

Usll

and,giventhesize ofthe

Usll

protein, it seemed doubtful thatthis entire sequence could be in contact with theprotein. Totestthelatteridea,

wedetermined the size of the RNA fragment protected from RNase T1 digestion by HSV-1 and HSV-2 RNA-binding

activity. The RNA probe was incubated with

HSV-1(F)-infected cell nuclear extract and HSV-2(G)-infected cell

nuclearextract,digested with RNase T1,and

electrophoret-icallyseparated in a nondenaturing polyacrylamide gel along with undigested probe-A.5m protein complex (Fig. 2B).

Binding complex bands (Fig. 2B; indicated by arrowheads) were excised from the gel, and RNA was purified and

subjected to electrophoresis on an 8% denaturing

urea-polyacrylamide gel along with intact probe and molecular

weightstandards. The results of RNase T1digestions (com-pareFig. 2B, lane 2, with lanes 3 and 4) indicated that HSV-1 and HSV-2 RNA-binding activities protected fragments of

identicalsize and that the fragments were 95 to 100 nt in size.

DISCUSSION

Our studies indicate that

Usll

encodes a protein which

participatesin and is necessary for sequence-specific binding to anin vitro RNA transcript of an HSV-1 DNA sequence.

Severalaspects of these results require further comment. The

Usll

openreading frame encodes a protein predicted to have a translated Mr of approximately 17,800.

The,

pre-dictedaminoacidcompositionis that of an extremely basic

proteinwith apredicted pl equal to 11.1(by the method of

Silleroand Ribeiro

[39]).

Itsstructure is unusual in that the

C-terminalhalf of the

protein

is

composed

of tandem reiter-ations of the sequence

Arg-X-Pro.

The number of these repeats varies(33,

42)

and

probably

accountsfor the differ-encein the sizes of

Usll

proteins

of different HSV-1 strains

(20). Wewere

perhaps

fortunate inourchoice of HSV-1 and HSV-2strainsfor

intertypic mapping

inasmuchas

HSV-1(F)

specifies

a

relatively

small

Usll

(20)

and

complexes

from

strains

specifying

a

larger protein might

notbe

distinguish-ablefromthose formed

by HSV-2(G).

The

relatively large

size of the

probe fragment protected

fromRNase

digestion by

the

binding complex (95

to 100

nt)

may suggestthatthe

binding complex

containsmorethana

single

moleculeof

Usll

and

perhaps

contains other

proteins.

Theobservation thatasmall

change

in the size of the

Usll

polypeptide produced

a

significant mobility

shift in the

complex, however,

suggests that

Usll

is

likely

to be the

major

protein

constituent ofthe

complex.

The

chromato-graphic

properties

of the

RNA-binding activity

on

Sephacryl

S300

(activity peaks

in the void

volume)

werecharacteristic

ofa

protein

much

larger

than that

predicted

from the size of

the

Usll

open

reading

frame.

However,

the nuclearextract

could be fractionated with ammonium sulfate such that the

RNA-binding

activity

ran as a

discrete,

lower-molecular-weight peak

in

Sephacryl

S300 columns

(work

in

progress).

These results and the observation that the

RNA-protein

complex

was

only

minimally

retarded relative to that of

unbound RNA suggest that

(i)

Usll

existsininfectedcells in

large

aggregateswhich

readily

dissociateintomonomers or

small multimers and

(ii)

the

U.11

containedin

RNA-protein

complexes

consists ofa

specific

but smallnumberof

mono-mers.

Thefunction ofthe

Usll

open

reading

frame inthe viral

reproductive cycle

isunknown. MacLeanetal.

(21)

reported

that the

protein product

ofthe

Usll

gene binds to DNA cellulose and is eluted

only

with a

high

salt

concentration,

consistent with a role as a

DNA-binding protein.

It has further been

suggested (21)

thatthe

Usll

protein

is the

Mr

21,000

to

22,000

-y

protein reported

to interact

specifically

with the HSVasequence(6). Thissequenceis locatedatthe terminiof the S and L components ofHSV DNA

(41)

and containsseveral

cis-acting

sites, including

those

required

for

cleavage

of concatemers of viral DNA into

unit-length

molecules and

packaging

ofthe viral genome

(reviewed

in

reference 34). Several lines of evidence suggestthat

Usll

doesnot

play

arole in

cleavage

and

packaging

anddoesnot,

in

fact,

bind viral DNA in vivo.

(i)

The

Usll

open

reading

framecan be deletedfrom the genomewith no apparentill effects for virus

growth

incell culture

(18, 23).

(ii)

Adetailed

study

of thefactors which bindtothe HSV-1asequence(4) showed no evidence of

binding

of

Usll

protein.

(iii)

Anti-body

localization

experiments

(21) show that

Us1l

protein

is

distributed

throughout

both the nucleus and thecytoplasmof infected cells and is concentrated in the nucleolus. This

distribution is more consistent with

affinity

for RNA than

with

affinity

for DNA inasmuch as viral DNA is excluded

from thenucleolus(30, 32)and incytoplasmispresentonly

in

capsids

(34). It is not, however, surprising that an

ex-tremely

basic protein such as

Usll

might bind

nonspecifi-cally

to DNA, or any other polyanion for that matter, in

vitro.

Thesecondarystructurespecificityof

Usll

binding might

have oneoftwogeneralcauses. It may be that the binding

site canbe recognized only insingle-stranded form, and in

nonbinding

RNAisomers this site existsatleast partlyina

stem-loop

structure.

Alternatively,

binding by

Usll

may

J.VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

HSV RNA-BINDING ACTIVITY 3469

require the presence of a particular secondary structure element not present in the nonbinding forms of the probe. It is formally possible that

Usll

recognizes some secondary RNA structure element(s) regardless of its sequence.

The probe to which

Usll

specifically bound in these experiments is complementary to the a47 and

Usll

mRNAs. The sequences protected by the RNA-binding activity are not present in any of the previously mapped transcription units, and sensitive RNAse protection assays on infected cell RNA failed to detect a corresponding transcript (work in progress). These observations suggest two mutually exclu-sive alternatives. First, the

U.11

RNA-binding activity may act as a negative regulator of RNA sequences to which it binds, and the probe sequences may be transcribed but fail to accumulate in significant amounts in cells which express

Usll.

To test this hypothesis, we must construct a virus in which the

Usll

RNA-binding activity is not expressed but in which the probe sequences and surrounding sequences are not grossly disrupted. None of the

Us11-

deletion mutants meet this requirement.

Second, the probe sequences used here may not be transcribed in vivo and are merely a substrate analog. While this is an attractive hypothesis, the observation that the 5' transcribed noncoding domain of the gene contains the target sequences of the gene product is in fact too much of a coincidence. To probe this hypothesis further, wehave RNA affinity studies for the

Usll

RNA-binding activity in prog-ress.

The function of the product of the

Usll

open reading frame is not known, but several possible functions which could result from binding of RNA late in infection have been excluded. Specifically, (i) host protein synthesis is decreased grossly late in infection bya viral function other than that of the vhs gene (encoded by the UL41 open reading frame), which degrades RNA upon entry of the virusintothe cell(8, 15, 31).

Usll

is not involved in the late degradation of host RNA inasmuch as the cells infected with the wild-type parent and the(x47-Usll deletion mutant R3631 accumulate equivalent amounts of cellular

(P-actin)

and viral(aTIF and glycoprotein C) mRNAs (data not shown). (ii) HSV encodes several transcripts, including those encoding the single-stranded DNA-binding protein (UL29) (9), the UL38 open reading frame (2), and possibly thymidine kinase (9), which terminate at more than one polyadenylation signal, and HSV-1 encodes or induces a factor which increases the efficiency with which at least one of these signals is used (26). That

Usll

is not a necessary component ofthis factor is suggested by the equivalent use ofthe UL29 and UL38 polyadenylation signals in HSV-1(F)- and R3631-infected cells (work in progress). (iii) Immunohistochemical studies (21) showed that

Usll

is associatedwith thenucleolus ofthe infected cell, suggesting a possible association with ribo-somes. HSV infection has a profound effect on ribosomal RNA metabolism (42), and a viral protein with sequence-specific affinity for ribosomal RNAcould play a role in the modulation of translation or turnover of specific mRNAs; components of a cellular system which differentially de-grades mRNAs have been found to be ribosome associated (37, 38). While no differences betweencellsinfectedwith the parent and R3631 mutant have been noted, experiments on this point are still in progress.

ACKNOWLEDGMENTS

These studies were aidedbyPublic Health Servicegrantsfromthe National Cancer Institute (CA47451) and the National Institute for

AllergyandInfectious Diseases(AI124009 and A11588-11). R.J.R. is apostdoctoral fellow of the American Cancer Society.

LITERATURE CITED

1. Ackermann, M., J. Chou, M. Sarmiento, R. A. Lerner, and B. Roizman. 1986. Identification by antibody toa synthetic peptide of aprotein specified by a diploid gene located in the terminal repeats of the L component of herpes simplex virus genome. J. Virol. 58:843-850.

2. Anderson, K.P., R. J. Frink, G. B. Devi, B. H. Gaylord, R. H. Costa,and E. K.Wagner. 1981. Detailed characterization of the mRNA mapping in theHindIll fragment K region of the herpes simplex virus type 1 genome. J. Virol. 37:1011-1027.

3. Chou, J., and B. Roizman. 1986. The terminal a sequence of the herpes simplex virus genome contains the promoter of a gene located inthe repeat sequences of the L component. J. Virol. 57:629-637.

4. Chou, J., and B. Roizman. 1989. Characterization of DNA sequence-common and sequence-specific proteins binding to cis-acting sites for cleavage of the terminal a sequence of the herpes simplexvirus 1 genome. J. Virol. 63:1059-1068. 5. Chou, J., and B. Roizman. 1990. The herpes simplex virus 1

gene forICP34.5, which maps in inverted repeats, is conserved in several limited-passage isolates but not in strain 17syn+. J. Virol.64:1014-1020.

6. Dalziel, R. G., and H. S. Marsden. 1984. Identification of two herpes simplex virus type 1-induced proteins which interact specifically with the a sequence of herpes simplex virus DNA. J. Gen. Virol. 65:1467-1475.

7. Ejercito, P.M., E. D. Kieff, and B. Roizman. 1968. Characteri-zation of herpes simplex virus strains differing in their effecton social behavior ofinfected cells. J. Gen. Virol. 2:357-364. 8. Fenwick, M. L. 1984. The effects of herpesviruses on cellular

macromolecular synthesis, p. 359-390. In H. Fraenkel-Conrat and R. R. Wagner (ed.), Comprehensive virology, vol. 19. PlenumPublishing Corp., New York.

9. Holland, L. E., R. M. Sandri-Goldin, A. L. Goldin, J. C. Glorioso, and M. Levine. 1984. Transcriptional and genetic analysis of the herpes simplex virus type 1 genome: coordinates 0.29 to 0.45. J. Virol. 49:947-959.

10. Honess, R. W., and B. Roizman. 1974. Regulationofherpesvirus macromolecular synthesis.I. Cascade regulationof the synthe-sis of three groups of viral proteins. J. Virol. 14:8-19. 11. Honess, R. W., and B. Roizman. 1975. Regulationofherpesvirus

macromolecular synthesis: sequential transitionofpolypeptide synthesis requires functional viral polypeptides. Proc. Natl. Acad. Sci. USA72:1276-1280.

12. Hubenthal-Voss, J., R. A. Houghten, L. Pereira, and B. Roiz-man. 1988. Mapping of functional and antigenic domains ofthe a4 protein of herpes simplex virus 1. J. Virol. 62:454-462. 13. Johnson, P. A., C. MacLean, H. S. Marsden,R.G. Dalziel, and

R. D. Everett. 1986. The product ofgeneUs11 of herpes simplex virus type 1 is expressed as a true late gene. J. Gen. Virol. 67:871-883.

14. Kristie, T. M., and B. Roizman. 1986. a4, the majorregulatory protein of herpes simplex virus 1, is stably and specifically associated with promoter-regulatory domains of a genesand of selected other viral genes. Proc. Natl. Acad. Sci. USA 83: 3218-3222.

15. Kwong, A. D., J. A. Kruper, and N. Frenkel. 1988. Herpes simplex virus virion host shutoff function. J. Virol. 62:912-921. 16. Lee, K. A. W., A. Bindereif, and M. R. Green. 1988. A small-scale procedure for preparation ofnuclear extracts that support efficient transcription and pre-mRNA splicing. Gene Anal. Tech. 5:22-31.

17. Leibold, E. A., and H. N. Munro. 1988. Cytoplasmic protein binds in vitro to a highly conserved sequencein the 5' untrans-lated region of ferritin heavy- and light-subunit mRNAs. Proc. Natl. Acad. Sci. USA85:2171-2175.

18. Longnecker, R., and B. Roizman. 1986.Generation ofan invert-ing herpes simplex virus 1 mutant lacking the L-S junction a sequences, an origin of DNA synthesis, and several genes including those specifying glycoprotein E andthea47gene. J.

VOL. 64, 1990

on November 10, 2019 by guest

http://jvi.asm.org/

3470 ROLLER AND ROIZMAN Virol.58:583-591.

19. Longnecker, R., and B. Roizman. 1987. Clustering of genes dispensablefor growth in cell culture in the small componentof the herpes simplexvirus 1genome. Science 236:573-576. 20. Lonsdale, D. M., S. M. Brown, J. H. Subak-Sharpe, K. G.

Warren, and H. Koprowski.1979. Thepolypeptide and theDNA restriction enzyme profiles ofspontaneous isolates of herpes simplex virus type 1fromexplantsof human trigeminal, supe-riorcervicaland vagus ganglia. J. Gen.Virol.43:151-171. 21. MacLean, C. A., F. J. Rixon, and H. S. Marsden. 1987. The

products of gene Usll of herpes simplex virus type 1 are DNA-binding and localize to the nucleoli ofinfected cells. J. Gen. Virol. 68:1921-1937.

22. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: alaboratory manual. ColdSpring Harbor Laboratory, ColdSpring Harbor, N.Y.

23. Mavromara-Nazos, P., M. Ackermann, and B. Roizman. 1986. Construction and properties ofaviableherpessimplex virus 1 recombinant lacking the codingsequences of thea47gene. J. Virol.60:807-812.

24. McGeoch, D. J., M. A. Dalrymple, A. J. Davison, A. Dolan, M.C. Frame, D. McNab, L. J. Perry, J. E.Scott, and P. Taylor. 1988. ThecompleteDNAsequenceofthelong uniqueregionin the genome of herpes simplex virus type 1. J. Gen. Virol. 69:1531-1574.

25. McGeoch, D. J., A. Dolan, S. Donald, and F. J. Rixon. 1986. Sequence determinationandgeneticcontentoftheshortunique region ofthe genome of herpes simplex virus type 1. J. Mol. Biol. 181:1-13.

26. McLauchlan, J., S. Simpson, and J. B. Clements. 1989. Herpes simplex virus induces a processing factor that stimulates poly(A) siteusage. Cell 59:1093-1105.

27. Meignier, B., R. Longnecker, P.Mavromara-Nazos,A. E. Sears, andB.Roizman. 1988. Virulence and establishment of latency by genetically engineered deletion mutants of herpes simplex virus1. Virology 162:251-254.

28. Nielsen, D. A., and D. J. Shapiro. 1986. Preparationof capped RNAtranscripts using T7 RNApolymerase.NucleicAcids Res. 14:5936.

29. Post, L. E., and B. Roizman. 1981. A generalized technique for deletion of specificgenesinlarge genomes: a gene 22 of herpes simplex virus 1isnotessentialforgrowth. Cell 25:227-232. 30. Puvion-Dutilleul, F., E. Pichard, P.Sheldrick,F. Arratric, and

E.Puvion. 1985.Appearance ofhost-specificnucleolar proteins in intranuclear 'dense bodies' following herpes simplex infec-tions. Eur. J. Cell Biol.39:458-468.

31. Read, G. S., and N. Frenkel. 1983. Herpes simplex virus

mutantsdefective in thevirion-associatedshutoff of host

poly-peptide synthesis andexhibiting abnormalsynthesis ofa (im-mediateearly)viralpolypeptides.J. Virol. 46:498-512. 32. Rixon, F. J., M. A. Atkinson, and J. Hay. 1983. Intranuclear

distribution of herpes simplex virus type 2 DNA synthesis: examination by lightandelectron microscopy. J. Gen. Virol. 64:2087-2092.

33. Rixon,F.J.,andD.J.McGeoch. 1984. A 3' co-terminalfamily ofmRNAs from the herpes simplexvirustype1 shortregion: twooverlapping readingframes encode unrelatedpolypeptides one of which has a highly reiterated amino acid sequence. NucleicAcids Res. 12:2473-2587.

34. Roizman,B.,andA. E. Sears. 1990.Herpes simplexviruses and theirreplication, p. 1795-1894. In B. N.Fields, D. M. Knipe, R. M.Chanock,M. S.Hirsch,J. L.Melnick,T.P.Monath,and B. Roizman (ed.), Fields' virology, 2nd ed. Raven Press, Publishers,NewYork.

35. Roizman, B., and P. G. Spear. 1968. Preparation of herpes simplexvirus ofhightiter.J. Virol. 2:83-84.

36. Roller,R.J.,A. L.McCormick,andB. Roizman. 1989. Cellular proteins specificallybindsingle-anddouble-stranded DNA and RNA from the initiation site of a transcript that crosses the originofDNAreplicationofherpes simplexvirus 1. Proc.Natl. Acad. Sci. USA 86:6518-6522.

37. Ross,J., and G. Kobs. 1986. H4 histonemessengerRNAdecay in cell-free extracts initiates at or near the 3' terminus and proceeds3' to 5'.J. Mol. Biol. 188:579-593.

38. Ross,J.,G.Kobs,G.Brewer,and S.W. Peltz. 1987.Properties of theexonucleaseactivitythatdegradesH4histone mRNA.J. Biol. Chem. 262:9374-9381.

39. Sillero, A., and J. M. Ribeiro. 1989. Isoelectric points of proteins: theoretical determination. Anal. Biochem. 179:319-325.

40. Vlazny, D. A., A. Kwong,and N. Frenkel. 1982. Site-specific cleavage/packagingofherpes simplexvirus DNA and the selec-tive maturation of nucleocapsids containing full-length viral DNA.Proc. Natl.Acad. Sci. USA 79:1423-1427.

41. Wadsworth, S., R. J. Jacob, and B. Roizman. 1975. Anatomy of herpes simplex virus DNA.II.Size, composition,and arrange-mentofinverted terminal repetitions. J. Virol.15:1487-1497. 42. Wagner, E. K., and B. Roizman. 1969. Ribonucleic acid

synthe-sis in cells infected with herpes simplex virus. I. Patterns of ribonucleic acid synthesis in productively infected cells. J. Virol.4:36-46.

43. Watson,R.J., and G. F. Vande Woude.1982.DNAsequenceof animmediate-earlygene (IEmRNA-5)ofherpes simplex virus type 1.NucleicAcidsRes.10:979-991.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/