Localization of cis-acting sequences in the latency-related promoter of bovine herpesvirus 1 which are regulated by neuronal cell type factors and immediate-early genes.

0022-538X/92/106099-08$02.00/0

CopyrightC1992, AmericanSociety for Microbiology

Localization of cis-Acting Sequences in the Latency-Related

Promoter

of Bovine Herpesvirus 1

Which Are Regulated by

Neuronal Cell Type Factors and Immediate-Early Genes

A. C. BRATANICH AND C. J. JONES*

Centerfor Biotechnology, Departmentof Veterinary Science, University of Nebraska,

Lincoln,

FairStreetatEast

Campus Loop,

Lincoln,

Nebraska 68583-0905

Received8June 1992/Accepted16 July 1992

Bovine herpesvirus 1 (BHV-1) establishes a latent infection in sensory ganglionic neurons of cattle. During alatentinfection, a single latency-related (LR) transcript is expressed. This observation suggested that DNA sequencesin the LR promoter are

positively

regulated by neuralcell type factors. The regulation of the LR gene wasexamined in neuralcellsaswellas nonneural cells intransient assays. A 258-bpXbaI-SphI fragment from the LR promoter cis activated the herpes simplex virus type 1thymidine kinase promoter in rat pheochromo-cytoma(PC-12) cells and differentiated human (HCNIA) neurons. In contrast, cis activation was not observed withrat(Rat-2) fibroblasts, undifferentiated HCNLA cells, or bovine turbinate cells. Treatment of PC-12 cells with nerve growth factor increased transcriptional activity of the XbaI-SphI fragment. Exonuclease HI footprintingexperimentssuggested that nuclearfactors bind to theXbaI-SphIfragment. Theimmediate-early

genes of BHV-1 trans activated the LRpromoter, and DNA sequences 5' to the XbaI-SphI fragment were necessary for maximal stimulation. These results imply that neural-cell-type-specific factors and BHV-1 immediate-early genes

positively

regulate LR gene expression.

Bovineherpesvirus 1(BHV-1)isasignificant viral patho-genof cattle andcancauserespiratory disease,abortions, or

occasionally encephalitis (for a review, see reference 31).

Likeall membersof thealphaherpesvirussubfamily, BHV-1 establishesalatentorpersistent infection in sensory

gangli-onicneurons of the infected host (1, 10, 21,22). Thevirus can persist ina latent state for the lifetime of the infected host or can periodically reactivate and cause extensive damage tothe host. In contrast to the 70to80viral genes expressedduringalyticinfection ofbovinecells, viral gene expression is severely impaired during a latent infection. One smallregionof the genomeistranscriptionallyactive in

latently infected neurons, and this region is designatedthe

latency-related (LR) gene (14, 19-21). Transcripts originat-ing from the LR gene accumulate in the nuclei ofsensory neurons which are latently infected (14, 19-21). The LR

transcript is also detected in lytically infected cells and accumulates late afterinfection(14).

The transcriptional promoterwhich regulates expression

of the LR gene is contained within a 980-bpPstI fragment (12). Inprimaryculturesof rabbit neurons, the LR promoter is 10-fold more active than the simian virus 40 early pro-moterand enhancer. Viral infection alsopositivelyregulates LR promoter activity. Consequently, it was hypothesized that cis-acting sequences within the LR promoter are re-sponsibleforneuronalcelltypeexpressionandtrans activa-tionbyviralorvirus-induced factors.

Inthisstudy, tworegions within theLR promoterwhich were necessary for efficient trans activation by BHV-1

immediate-early (IE)geneswereidentified.A258-bp

XbaI-SphI fragmentcis activated theherpes simplexvirus type 1

(HSV-1) thymidinekinase(TK)promoterincells of neuronal

originbutnotother cell types.However,thisfragmentdoes nothave promoteractivity andwas notnecessaryfortrans

* Correspondingauthor.

activation byBHV-1 IE genes. Ifneuronal cellsweretreated withnervegrowth factor(NGF),transcriptional activitywas stimulated slightly. IftheXbaI-SphI fragmentwas digested withAIuI,cisactivation of theTKpromoterwasobserved in neuronal cells only after treatment with NGF. When the

XbaI-SphI fragment was incubated with nuclear extracts from rat cells, protection from exonuclease III (ExoIII) digestionwasobserved,suggestingthat nuclear factorsbind sequences in the XbaI-SphI fragment. These results indi-cated that distinctregionsof the LR promoterarenecessary for neuronal cell typeexpressionandtransactivationbyIE genes.

MATERIALSANDMETHODS

Cells. Bovine turbinate and Rat-2 cellsweremaintainedas described previously (11, 12). PC-12 cells(rat

pheochromo-cytoma[9])wereobtained from the American

Type

Culture

Collection and maintained in RPMI media

supplemented

with 8% horse serum and 4% fetal bovine serum. HCN1A cells (human cortical neurons [23]) were maintained in Earle'smodifiedEaglemediumsupplementedwith15%fetal bovine serum. Todifferentiate HCN1Acells,cultureswere treated with 0.5 mM 3-isobutylmethylxanthine

(IBMX),

0.5 mMdibutyrylcyclicAMP(cAMP),and 25 ng of NGF per ml for 6days(23). During differentiation, cellswerefedevery3

days. Rat-2, BT,and PC-12 cellsweretreatedwith 25 ng of NGF per ml for the indicated times.

Cell transfection. Transfection of BT, Rat-2, and PC-12 cellswascarriedoutbycalcium phosphate

precipitation

as describedpreviously (11,12).HCN1A cellsweretransfected

bythe Polybrenemethod(18).

CAT assays. Forty hours

posttransfection,

cells were washed in phosphate-buffered saline and a cell-free lysate waspreparedby three freeze-thawcyclesin 0.25MTris

(pH

7.8). Chloramphenicol

acetyltransferase

(CAT)

enzymatic

activitywasmeasuredasdescribed

previously

(11,

12).

After

6099

on November 9, 2019 by guest

http://jvi.asm.org/

6100 BRATANICH AND JONES

thin-layer chromatography (TLC),

theamountof

[14C]chlo-ramphenicol (CM), acetylated

or

unacetylated,

was mea-sured withanAmbis

radioanalytical

device.

Preparation

of nuclear extracts. Nuclear extracts were

prepared

from mammalian cells

by

the

procedure

ofDignam

etal.

(7).

Proteinconcentrations of the variousextractswere measured

(4),

and theextractswere

subsequently

frozenat -1200C.

Plasmid constructs. Plasmids

pO.9Scat/l, pO.95cat/6,

pAXhoI, and pO.SXba/47 were described previously (12). Plasmids

pSV2cat,

pAlOcat2,

and

pSVOcat

were obtained from BruceHoward. Toconstruct

pAXP/1,

the

unique

XhoI site in

pAXhoI

was convertedto a

BglII

site

by

addition of

synthetic BglII

linkers and the

531-bp XhoI-HindIII

frag-ment

(nucleotides [nt]

451to

981)

wasreleased

by digestion

with

BglII-HindIII

and

subsequently

inserted into a

pAlOcat2

vector

digested

with

BglII

andHindIII. Prior to

cloning,

the

pAlOcat2

vector was

purified

in a

polyacryl-amide

gel

to eliminate simian virus 40

early

promoter se-quences.The

plasmid pAXS/1

wasconstructed

by digesting

pAXP/1

with

SphI

and

HindIII,

filling

in the ends with Klenow

fragment,

and

subsequently ligating

with T4ligase.

Forasummaryofthese constructs, see

Fig.

1.

For constructionof enhancer

plasmids,

aCATconstruct

containing

the HSV-1 TK promoter was used

(pBLcat/4

[26]).

Plasmid

pO.95cat/1

was

digested

with

XbaI,

which releasednt523to981as

anXbaI

fragment.

Thiswas

possible

since the SmaI-PstI sequences of the

pUC-19 polylinker,

including

an XbaI

site,

are

adjacent

to nt 981 of the LR promoter.nt523to981 of the LR promoterwerecloned into

pBLcat/4

attheXbaI

site,

andaclone withnt981

adjacent

to theTKpromoterwasselected

by

restriction

mapping.

This constructwas

digested

with

SphI,

whichreleased the

XbaI-SphI

site of the LR promoter

(523

to

781)

by

virtue ofan

SphI

site in the

polylinker

of

pBLcat/4.

The

resulting

linear-ized

plasmid

was recircularized with T4

ligase

and

desig-nated

pBL/Sph-.

Thereleased

SphI

fragment (XbaI-SphI)

wascloned into the

SphI

site of

pBLcat/4,

and the

resulting

constructwas

designated pBL/Sph+.

Theorientationof the

fragment

was determined

by digestion

withXbaI. The

Hin-dIII-XbaI

fragment

of the LR promoter

(1

to

523)

wascloned

into the HindIII-XbaI sites of

pBLcat/4

and

designated

pBL/H-Xba.

Forasummaryof these constructs, see

Fig.

3. All DNA

fragments

were

purified

in5%

polyacrylamide

gels

and electroeluted. Plasmids were

purified

from bacteriaby

alkaline

lysis

followed

by

two CsCl

density centrifugations

(25).

Plasmids

containing

the BHV-1 IEtul

region (c601)

and E2.6wereobtained from M.

Schwyzer (28, 29).

To prepare

a construct that

expressed

only IE4.2,

c601 was

digested

with

SalI

plus HindIII,

the ends were made blunt with

Klenow

fragment,

the

large fragment

was

purified by

agarose

gel electrophoresis,

and the

fragment

was

ligated

with T4 DNA

ligase.

Forasummaryofthese constructs,see

Fig.

2.

Binding

assays for ExoIlH

digestion.

The

XbaI-SphI

frag-mentwaslabeledatits 5'

phosphate

ends

by incubating

with

polynucleotide

kinase and

[-y-32P]ATP.

The

fragment

was

subsequently digested

with

AluI,

and the individual

frag-ments were

purified

in

polyacrylamide

gels as described above.

Binding

volumes for

ExoIlI digestion

were50 ,ul. The reaction mixtures contained thefollowing components: 0.5 to 1 ngofend-labeled DNA

(10,000

to 20,000 cpm), 45mM

KCl,

25 mM HEPES

(N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic

acid) (pH 7.8),

0.1 mM EDTA, 0.5 mM

dithiothreitol,

3 ,ug of

poly(dI-dC),

and 15 p,g of nuclear extract.After the reactionwasincubated for 20 minat30°C,

MgCl2wasaddedto afinalconcentration of5 mM and 375 U ofExolII wasalso added. This is theamountofExoIII that yielded little or no background when the fragments were incubated withjust bovineserumalbumin (datanot

shown).

Reaction mixtures were incubated foranadditional 20 minat 30°C, and reactions weresubsequently terminatedby addi-tion ofphenol-CHCI3-isoamyl alcohol (50:49:1). After etha-nol precipitation, the products were separated on a 6%

denaturing polyacrylamidegel.

For competition footprinting assays with the

respective

oligonucleotides, 20or200ng of the double-stranded

oligo-nucleotide was incubated with the nuclear extract mixture for 20minat 37°C. The labeledfragmentwas

subsequently

added, and the reaction mixturewasincubatedasdescribed above.

PC-12, Rat-2, or HCN1A cells were transfectedwith the designatedplasmids (10 ,ug). Twenty-four hours posttrans-fection, some cultures of Rat-2 cells or PC-12 cells were treated with 25 ng of NGF per ml since thisconcentration wasshowntoinduce differentiation of PC-12 cells as well as induce early-response genes(5, 9,24). For differentiation of HCN1A cells, cultures were treated with NGF (25 ng/ml),

IBMX (0.5 mM), and forskolin (0.5 mM) for 6 days prior to transfection (23). At 40 hposttransfection,totalcelllysates wereprepared and CAT enzymatic activity was measuredin the presence of

[14C]CM

(40 p.g of protein, 1 h, 37°C). For HCN1Acells, extractsderived from an equivalent number of cells (2 x 104) were used to measure CATactivity (4h,

37°C). The various forms of CM were separated by TLCand

quantified.

The 21-bp oligonucleotide which contains three AP-1 or tetradecanoyl phorbol acetate-responsive elements (TRE)

is 5'-TGAGTCATGAGTCATGAGTCA-3' (2). The 39-bp

oligonucleotidewhichcontains three cAMP-responsive ele-ments (CRE)is5'-GATCTGACGTCATGACTGACGTCAT GACTGACGTCATCA-3' (17). The 36-bp oligonucleotide whichcontains three octamer (Oct)-binding sites is 5'-GAT CATGCAAATGATCATGCAAATGATCATGCAAAT-3' (8).

RESULTS

Localization of the minimal LR promoter. To localize

regions of the LR promoter necessary for transcriptional

activity, restriction fragments from the LR promoter were insertedatthe 5' terminus ofaplasmid carrying a promoter but not CAT and transcriptional activity was measured in bovineturbinate(BT) cells,ratpheochromocytoma (PC-12)

cells, or tat (Rat-2) fibroblasts. Previous results indicated that the LR promoterwas an efficient promoter in sensory neuronsbutnotother cell types(12).However, these studies did not localize elements inthe LR promoter which were necessary for efficienttranscription in neural cells. As

pre-viously

demonstrated, LR promoteractivity had strict ori-entationpreferences in allcell lines tested since pO.9Scat/1 wasanefficientpromoterbut pO.9Scat/6 was not (Fig. 1). In BTand Rat-2cells, the simianvirus 40 early promoter and enhancer(pSV2cat)wasfivetosixtimes more efficient as a promoter thanwas pO.95cat/1. In contrast, pO.9Scat/1 had

higherpromoteractivityin rat neuroblastoma (PC-12) cells. A construct containing the XhoI-PstI fragment (nt 451 to

981; pAXP/1)had promoteractivity similartothat in pAXhol inBT, PC12,and Rat-2cells. If theXhoI-XbaI fragment was removed (nt451to523;pO.SXba/47),promoteractivity was abolished. When sequences between SphI-PstI (nt 812 to

981)

weredeleted,promoteractivity was eliminated. These J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/

A. pO.95cat/ 1

r-r-l- - In a'4%:

Or- N - at0 Co

_t) vX'uo'0 r- 0%

A Ssp

H/P X A X Xb AP

11

I II

k k

I C

_!v&&&&x_I1I1I I CAT

BT 12.5 ± 4.5

%acetylated CM Rat-2 10.2 + 4.1

PC-12 64.2 + 19.2

B. pO.95cat/6 I 11I1X'NNN CAT 0.2 + 0.1

C. p&Xhol D. pAXP/ 1 E. pO.5Xba/47 F. pAXS/1

CAT

11111 I CAT

CAT

11111 I CAT

59.2 + 12.3 54.4 29.8 0.5 + 0.3 0.7 + 0.3

6.3 + 1.9

7.8 + 3.5 0.2 + 0.1 0.3 ± 0.1

39.5 + 19.2

33.6 + 13.5

6.4 + 1.4

0.2 + 0.1

uRnN

CAT 65.8 15.7 _68.4 + 19.2 49.7 + 13.4

H. pSVOcat CAT 0.2 ± 0.1 _0.2 + 0.1 0.5 + 0.2

FIG. 1. Localization of LRpromoteractivity. Cellsweretransfected with various plasmids (10 p.g).At40hposttransfection, total cell

lysateswereprepared and CATactivitywasmeasured in thepresenceof[14C]CM(40 ,ug of protein, 1 h, 37°C). The various forms ofCMwere

separated by TLC andquantified. Restrictionenzymesitesare:PstI (P), XhoI(X), AluI (A), XbaI (Xb), SspI (Ssp), SphI(S), and HindIll (H). The arrow indicates the direction of LR transcription and the approximatestart site. Nucleotide numbering is from the published

sequence(14). Resultsaretheaveragesfromthreeindependentexperiments.

results indicated that inbovine cells andratcells (neuralor

nonneural), the minimal LR promoterwas localized to the XhoI-PstIfragment (nt 451 to981).

trans Activation of LR promoter by BHV-1 IE genes. To analyze the ability of BHV-1 IE genes to regulate LR promoter activity, cotransfection experiments were

per-formed with BT cells. Previous experiments demonstrated that LR promoter activity and steady-state levels of LR RNAaccumulateatlatetimesafterinfection, suggesting that viral genes or virus-induced factors were responsible for

activation (12, 14). BT cells were cotransfected with the

respectiveLRpromoter constructsandaplasmidcontaining one ormore oftheBHV-1 IEgenes (for a summaryof the

BHV-1 IEgenesused inthis study, see Fig. 2). Aplasmid

containing IEtul(c601)orjust IEtul/4.2positively regulated

the intact LR promoter (pO.95cat/1) sixfold. If the XhoI fragment(nt 307to452;pAXhoI)orPstI-XhoIfragment (nt

1-452; pAXP/1) was deleted, trans activation by c601 and IEtul/4.2 was reduced almost threefold. In contrast, E2.6

trans activatedpAXhoI as efficiently asthe intact LR

pro-moter. However, transactivation ofpAXP/l by E/2.6 was

reducedmorethantwofold.These resultsdemonstratedthat

theXhoI fragment of the LR promoter was necessary for

trans activationby IEtul/4.2 and thatsequenceswithin the PstI-XbaI region were necessary for trans activation by E/2.6.

cis Activation of the TK promoterby LR promoter frag-ments. To determine whether regions of the LR promoter have transcriptional enhancer activity, DNA fragments whichspannedthe LRpromoterwereinsertedadjacenttoan

HSV-1 TK promoter construct (pBLcat/4) and transcrip-tional activity was measured in PC-12, BT, or Rat-2 cells.

The PC-12 (rat pheochromocytoma) cell line has features whicharesimilarto thoseofneurons, canbedifferentiated

into neuronlike cells by numerous agents, and has widely beenusedtostudyneural-cell-type-specificevents(5, 9, 24).

Rat-2 cells are a fibroblast cell line (11) and served as a

nonneural cell type to test for cell-type-specific cis-acting

sequences. InPC-12 cells,theXbaI-SphI fragment oftheLR

promoter (pBL,/Sph+) enhanced TK promoter activity 15-fold (Fig. 3). Constructs containing either the SphI-PstI fragment (nt 812 to 981; pBL/Sph-) or the HindIII-XbaI fragment (nt 1 to 523; pBL/H-Xba) had promoter activity thatwas at least threefold higher than that ofpBLcat/4 in PC-12cells. In Rat-2cells, however, promoteractivitiesof pBLVSph+ andpBVSph-werelowerthan that ofpBLcat/4.

In BT cells, promoter activity of pBVL_Sph+ was slightly

higher than that of pBLcat/4. These results demonstrated thatanXbaI-SphIfragment cis activated the TKpromoterin rat pheochromocytoma (PC-12) cells but not in nonneural (Rat-2orBT)cells.

Effect ofNGF on LR enhancer activity. To examine the effect NGF has on LR enhancer activity, cellswere trans-fected with the various enhancer constructs, in thepresence of NGF, and promoter activity was measured. In PC-12 cells, promoter activity of pBLVSph+ was slightly higher

after cells were treated with NGF (Table 1). Promoter

activityof the other enhancerconstructswasnotaltered in Rat-2 or PC-12 cells by NGF treatment. These results demonstrated that NGF stimulated enhanceractivityof the XbaI-SphI fragmentinPC-12 cells butnot Rat-2 cells.

To further explore the effects ofNGF on LR enhancer activity, the humancorticalneuron cellline, HCN1A,was

transfected with the variousconstructs and promoter activ-itywas measured inundifferentiated ordifferentiated cells.

HCN1A cellswereobtained froman individual with

unilat-eral megalencephaly (23). These cells express neuron-spe-cific but notglia-specific markers and canbe differentiated with dibutyryl cAMP, IBMX, and NGF. After differentia-tion, cells displaya mature neuronal morphologywith nu-merous long, extensively branched processes with spines and varicosities(lla, 23). InundifferentiatedHCN1Acells,

0.2 + 0.1

G. pSV2cat

0.3 + 0.1

on November 9, 2019 by guest

http://jvi.asm.org/

[image:3.612.62.549.78.294.2]

6102 BRATANICH AND JONES

A.

0

-3

0

la

la

0s

* IEtul o IEtul/4.2

El E2.6

pO.95cat/1 pAXhol pAXP/1

Map Units

C

Transript

*Positions

D. Plasmid

*Desigwnation

c601 IEtul/4.2

| <

L

0-|

S-

-|

g

0 0.2 0.4 0.6 0.8 1.0

- E2.6 - e2 IE4.2 el ORI e1 e2 e3

2 _ I E2.9

-LR

7U)u C ) cn. cxuZ=

E z n f) (f -zz z(f

1000 bp

[image:4.612.137.472.81.484.2]

E/2.6

FIG. 2. transActivationofthe LR promoterby BHV-1genes.(A)BTcellsweretransfected with 10pLgof therespectivepromoter-CAT plasmids and 8 ,ugofaplasmid containingBHV-1 IE genes. At48hposttransfection,CATenzymatic activitywasmeasured. Valuesare

percentsacetylated CM insamplescotransfected withLRpromoterconstructs andthedesignatedIEplasmidsdividedbycontrolvalues (withoutanyIEconstruct).Valuesareaveragesforfiveindependentexperiments. (B) Diagramof BHV-1genome.(C)Positions of IEgenes. IE4.2 isthe4.2-kbRNAtranscript derivedfromlEtul. IE2.9 is the 2.9-kbtranscript. Solid linesin thetranscript positionmaprepresentexons (elore2).Dashed lines indicate thepositionsof intronsandarrowsindicatethe direction oftranscription. (D)IEcontainingplasmidsused inthis study. RestrictionmapofthelEtulregion is derivedfromreference 29. E/2.6 isanearly2.6-kbtranscriptwhich encodesaprotein identicaltoIE/2.9(28).

pBLcat/4 had higher activity compared with those of LR enhancer constructs (Table 1). In differentiated HCN1A

cells, pBL/Sph+ had10-fold-higher promoter activity than

pBLcat/4. Thus, in differentiated human neurons, the TK promoter was cisactivatedby theXbaI-SphIfragment.

Localization ofelements intheXbaI-SphI fragmentwhich are necessaryfor enhancer

activity.

Tobegin tounderstand whichregions of theXbaI-SphI fragmentwerenecessary to cis activate the TK promoter in PC-12 cells, the fragment wasdigestedwithAluIandindividualfragments were tested

for enhancer activity. The results indicated that the

AluI-SphI fragment

(pBLSph2A)

stimulated TK promoter

activ-ity approximately twofold in PC-12 cells as well as Rat-2 cellscomparedwithpBLcat/4 (Fig.4). TheXbaI-AluI

frag-ment

(pBL/XballC)

hadlittle,if any,effectonTK promoter

activity in PC-12 or Rat-2 cells. When PC-12 cells were treated with NGF, promoter activities of pBL/Sph2A and

pBL/XbalC

werestimulatedatleasttwofoldcomparedwith thosefor normal cells. If Rat-2 cellsweretreatedwithNGF,

promoteractivityof

pBL/SphlA

andpBLcat/4wasreduced

approximately twofold compared with those for normal cells. Individual AluI

fragments

werenotefficientenhancer elements inPC-12cellscompared with theintactXbaI-SphI fragment, suggesting that sequences neartheAluI site are necessary for transcriptional enhancer activity in PC-12 cells.

Binding of cellular factors to the LRenhancer fragment.

ExoIlI

footprinting assays were conducted to determine

J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/

A

Plasnid 1 2 3 4

7 '

;j40 sige

6

_1

Ac-CMl(%) "2 "

B

_{1. plilcal/4}

2.

pflB,JSph-3. pli11 1-XAba

S.

[image:5.612.60.295.70.326.2]

4. plIf/Sph+

FIG. 3. Localization of DN transcriptional enhancer activ

variousplasmids (10 tg). At4

wereprepared and CATactiv

[14C]CM (40 ,ugofprotein,1h, separatedby TLC, percents

thesevalues aregiven. The n

plasmids presentedinpanelB thisstudy. Restriction sitesa

BamHI(B), and XhoI(X).

whether cellular factors b( ExoIII is a 3' to 5' exc

beledXbaI-AluI fragment. Addition of theCRE oligonucle-otidehadnoeffectontheExoIII protectionpatternin Rat-2

PC-12 .5T-2 or PC-12 cells. The Oct oligonucleotide inhibited binding

X1 23 4 1 2 3 4 slightly to regions A and B in PC-12 and Rat-2 cells. The

TREoligonucleotide also reduced binding slightlytoregion CinnuclearextractspreparedfromPC-12cells butnotRat-2 cells.An examinationofthesequence(Fig. 6) didnotreveal

aTREinregion Cor an Oct-binding site in region AorB,

suggesting that the inhibition observed was nonspecific or

thatfactors that bindtosequencesintheXbaI-AluIfragment

*o ** ,* also haveanaffinity foraTREconsensusorOctconsensus. WhentheAluI-SphI fragmentwasincubated with nuclear

extractsprepared fromPC-12orRat-2cells, protectionfrom

:hX-Exoll digestion could be observed in three separate

do-mains (Fig. 5 and 6). NGF-inducible binding to the entire

IIsxI) _7- fragmentwasobserved forPC-12and Rat-2cells.However,

s'.Z.s2-7.E1 CAT qualitative differences inExoIIIprotectionpatterns andthe

protectionpatternsobtained withnuclear extractsprepared

S I'llNb from untreated cellswerenotobvious. WhentheAluI-SphI

.i.,

CA7 fragmentwasincubated with nuclearextractspreparedfrom x xXh

PC-12

cells, subtle differences in

ExoIII

protection profiles liS,TKwereobservedinregionsEand Fcomparedwith results for

_ ; CAT Rat-2cells (Fig. 5;denotedby asterisks). Even after treat-Xb s ment with NGF, the novel protection patternwas detectedin

K

t-ii-:}i5- regionF.WithinregionE,twoOct-likeelementsareflanked

by putativeTATA boxes(region F)andaCCAATbox(Fig. [Asequencesin theLRpromoterwith 6). When an oligonucleotide containing a consensus Oct-rity. (A) Cellswere transfected with binding sitewasusedasacompetitor in footprinting

exper-0 h posttransfection,total celllysates iments,binding toregion Ewasdiminished comparedwith ritywasmeasuredinthepresenceof ExoIllI protectionpatterns in control samples. In Rat-2cells,

a37C)yThe variousCformseofiCM

were the Oct

oligonucleotide

inhibited

binding

to two Oct-like

aumbers

above the lanes referto the elements in region E as wellas surrounding sequences. In (B)Diagram of the plasmids usedin

PC-12

cells, inhibition of binding to region E by the Oct re

HindiII

(H),SphI (S), XbaI(X!),) oligonucleotidewas confined to the twooverlappingOct-like

elements. Oligonucleotides containinga consensusTRE or

CREdidnotalter the ExoIIIprotectionpattern inregionD, E, orF intheAluI-SphI fragment. In summary,binding of nuclear factorstotheXbaI-SphI fragmentwascomplex,and

)und to the XbaI-SphI fragment. subtle differences in

ExoIII

protection profiles were

ob-nuclease which has been used servedforPC-12cells Rat-2cells.

successauiiy

Lo

anaIlyze

iteractuons

uetween

reguiatory

ele-ments and DNAbinding proteins in crude nuclear extracts (30). When nuclearextracts fromPC-12orRat-2cellswere

incubated with the XbaI-AluIfragment,atleastthree distinct regionswereprotected(Fig. 5;forasummaryofbinding,see

Fig. 6). If the XbaI-AluI fragment was incubated with nuclear extracts treated with NGF, enhanced binding was

observedin PC-12cellsbutnotinRat-2 cells. Oligonucleo-tidescontainingconsensussites forOct,CRE, orTREwere

incubated with nuclearextracts (20min) priortoaddition of the radiolabeled XbaI-AluI fragment to determine whether these respective transcription factors interact with the

la-DISCUSSION

In this study,we identified a258-bpXbaI-SphI fragment

from the LR promoter which cis activated the HSV-1 TK promoter in cells of neuronalorigin. NGF stimulated tran-scription in neuronal cells but not other cell types. The XbaI-SphI fragment did not have promoter activityin any

cell type tested, suggesting that other regions of the LR promoterwere requiredfor promoteractivity. ExoIII foot-printing experiments indicated that nuclear factors

inter-TABLE 1. Effect of NGF on LR enhanceractivity'

Mean(SD)forcell type:

Plasmid PC-12 PC-12treated Rat-2 treated HCN1A

with NGF Rat-2 with NGF Undifferentiated Differentiated

pBLcat/4 1.0 ± 0.3 1.0+ 0.4 1.0+ 0.2 1.0 ± 0.3 1.0± 0.3 1.0± 0.5

pBLVSph- 3.2+ 1.8 2.6 ± 1.3 0.2 ± 0.1 0.3 + 0.2 0.2+ 0.1 1.2± 0.8

pBL/Sph+ 17.2 ± 4.8 23.1 ± 3.1 0.6+0.4 0.2+ 0.3 0.5 ± 0.2 10.5± 2.9

pBL/H-XbaI 2.6 ± 0.8 2.2 ± 0.9 0.4± 0.1 0.5 ± 0.2 0.9± 0.3 1.2 + 0.4

aValuesgivenare a

comparison

ofpromoteractivityintherespectiveenhancerconstructs tothat inpBLcat/4.Each valuerepresentsthe averagefromatleast

threeindependentexperiments.Fordetails,seeMaterialsand Methods.

on November 9, 2019 by guest

http://jvi.asm.org/

[image:5.612.54.554.619.704.2]

6104 BRATANICH AND JONES

pBLcat/4

HSXbB- TK

CAT _1.0±0.3

Rat-2

1.0+0.4 1.0 + 0.3

Rat-2

NGF 1.0 + 0.3

:b S

ITK

P.;y.;.Y;m; CAT 17.2 + 2.8

A S

I I

eTK.i.

_CAT

2.6 + 0.6

Xb A

I I TK

[image:6.612.82.547.84.236.2]

CAT 1.4 + 0.5

FIG. 4. Enhanceractivity ofAluI subclones derivedfromXbaI-SphI fragment.Therespectivecellsweretransfected with variousplasmids (10 iLg).In somesamples,NGF(25ng/ml)wasaddedat16 hposttransfection.At 40 hposttransfection,total celllysateswerepreparedand CATactivitywasmeasured in thepresenceof[14C]CM (40Fgofprotein,1h, 37°C).The various forms of CMwereseparated byTLC and quantified.

acted with theXbaI-SphI fragment, and subtle differences were observed for PC-12 cellsversus Rat-2 cells. LR pro-moter sequences5'totheXbaI-SphI fragmentwererequired for trans activation by BHV-1 IE genes. These data sug-gestedthat atranscriptionalenhancer in the LRpromoteris active in neuronal cells and that other sequencesnearthe 5' terminusof the promoterwere transactivatedbyBHV-1IE genes.

XbaI-AluI

Rat-2 PC-12

<L_JLJ -W-L

C.,_ xt

B + .. 0 ..

*v-B

..

TT * ,.

..I

c .l 'r

U

**t41.

Alul-SphI

Rat-2 PC-12

*.'. ... liE

t**!fW[1F*

4 ,.. m.

: ;t*

Xp_ _l

3r

e

*:

tA*N

o lp D

FIG. 5. Analysis of nuclear proteins which interact with XbaI-SphI fragment by ExoIll digestion. TherespectiveAluI fragments

wereincubated with 151Lg of bovineserumalbumin (BSA), nuclear

extractfrom Rat-2orPC-12 cells (U), ornuclearextractprepared

from cellstreated with NGF(25 ng/ml, 24 h[NGF]). For

competi-tion experiments, synthetic oligonucleotides (20 or 200 ng)were

incubated for 20 min with nuclear extracts prepared from cells treated withNGF and the radioactive fragmentwassubsequently

added.ExoIII footprintingwasconductedasdescribed inMaterials

andMethods.

LR promoter activity and steady-statelevels of LR-spe-cific mRNA increaseduringalyticinfection,

suggesting

that aviral geneor avirus-induced factormediatesthis process

(12, 14). This study indicated that IEtul stimulated LR promoteractivity(Fig. 2). Sequenceswithin theXhoI-XhoI

fragmentwere necessary fortrans activation by

IEtul/4.2.

TheXhoI-XhoIfragment isanegative regulatoryelement in bovine cells, suggesting that other factors, in addition to IEtul/4.2, mediate LR transcription by interacting with

DNA sequencesin thefragment.ThePstI-XbaIfragment

(nt

1 to 307) was necessary for trans activation by E/2.6,

implying thattransactivationbyIE geneswasmediatedby at least two independent DNA sequences. These results suggested thatIEtul products coordinate LR gene expres-sion duringalyticinfectioncycleand that neuronal factors directhighlevels ofexpression duringalatentinfection.

Attempts to localize aLR promoterfragmentwhichwas active in PC-12 cells but not BT or Rat-2 cells were not successful, suggestingthatrequirementsfor basal promoter

activitywere similar for all cell types. In PC-12 cells and differentiated human

(HCN1A)

neurons,the

XbaI-SphI

frag-mentincreased TK promoteractivitymorethan10-fold. The

SphI-PstI fragment(nt812to981;

pBL/SphI-)

and

HindIIl-XbaIfragment(nt1to523;pBLIH-Xba)also increased TK promoter activitythree-to fourfold in PC-12 cells (Fig. 3).

However, these fragments did not cis activate the TK promoter in BT cells, Rat-2 cells, or undifferentiated HCN1A cells. The finding that treatment of HCN1A and

PC-12 cells with NGF led to higher promoter activity of

pBL/Sph+

butnotpBLcat/4 strengthensthehypothesisthat

cell-type-dependent interactions were crucial for cis-acting

functions of the

XbaI-SphI

fragmentsinceNGFis critical for neurondifferentiation andmaintenanceofneuronphenotype (forareview, seereference15).Insummary,theXbaI-SphI fragment in the LR promoter was a strong transcriptional

enhancer in cells of neuronaloriginwhen linkedto theTK promoter.Thefindingthat otherregionsof the LR promoter have weak enhancer activity in PC-12 cells implied that

expression of the LR gene in neural cells is not entirely dependenton onecis-actingsequence.

Several discrete regions within the XbaI-SphI fragment wereboundby nuclear factorsasjudged by

ExolII

footprint-xi

pBL/Sph+ t

pBL/Sph2A

pBL/Xba1C

0.6 + 0.4

1.9 + 0.5

1.2 + 0.5

23.1 + 0.2

6.5 + 0.9

4.1 + 0.6

0.5 + 0. 1

0.9 + 0.2

0.9 + 0.3

A

F; - -S -} -s a | [;;@;.;.;1

J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/

[image:6.612.63.302.407.626.2]

A G

I" V." r

LR

transcript

SaII

A B

Xbal

TGTCTRORCO GGGCTTTGTG TTTTTCCCRC CCCTOCCCTT GCCCCTGGCC GCCTGGCCGG 580

C NGF-IC LAT AP-2

_ ~~~~~~~~~~~~Alul

CCGGCTRRRG TATRGGCCRG RCCRRRCCCC CCGCRGCRGC TCTGCTCGCT GGCGRGRCCC 610

D NGF-IC E

CRRRGCGCCR RTTGTGGCRT RTTTCCGGGT CTGRCCGCGC CTGCRCTGC CTTTGGCRTT 700

F ICAAT I OCT OCT

RRTRTTTCRG flGTTTRRTRTT6CRGRCRGR CRRRRRGCCR GRTRRTTRCR RRGTRTTTGT 760

TATA TATA

TTTTRTTGRT TGCGCRTGCG

Sphu

FIG. 6. Summary ofExoIII protection experiments from LR enhancer region. (A) Restriction map of the LR promoter region and approximatestartsiteofLRtranscription. Regions oftheLR promoter which were necessary for trans activation by E/2.6 and IE/4.2 are indicated. Restriction enzymesites within the LR promoterare PstI(P), HindIII (H),XhoI(X),XbaI(Xb), AluI (A), SspI (Ssp), and SphI (S). (B)DNA sequenceofthe LR promoterfrom the coding strand of the LR transcript was derived from the published sequence (14). The lettersA to Fabove thesequencecorrespond to regions protected from ExoIII digestion after incubation with nuclear extractsprepared from PC-12 cells treatedwithNGF. Shownis theposition oftheNGF 1Cbinding site (NGF-1C;GCGGGGGCG [5]). Both NGF-1C sites contain sevenof ninematches andarepresent onthe complementary DNA strand. The AP-2 site matches the consensus (GCCNNNGGC[27]). Positions of the Oct-like elements haveanarrowin thedirectionof the consensus, are located on the complementary DNA strand, andcontain

twomismatches fromanOct-1consensus(ATGCAAAT; forareview,seereference8).Aregionof theHSV-1latency-associated transcript promoterwhich is within theneuron-specific enhancer (LAT [3])has 9of11basesmatchingbutis 3'to5'with respect totheHSV-1 sequence. Positions of the TATA-like elementsandthe CAAT boxaredenotedby the boxes. Regions ofthe LR promoterwhichareregulatedby BHV-1 IE genes arealsodepicted.

ingexperiments (Fig. 5). However, onlyminordifferencesin ExoIII protection patterns were observed when the XbaI-SphI fragment was incubated with nuclear extracts from PC-12cellsversusRat-2cells. InPC-12 cells, NGFinduces a family ofgenes which are designated the early-response

genes(5).Amember of thisfamily,NGF-1C, isa

transcrip-tional activator that specifically binds the sequence

GCGGGGGCG (5). Within the XbaI-SphI fragment, two regions are present with seven of nine bases matching

NGF-1C(Fig. 6).AregionwhichcontainsaconsensusAP-2 bindingsitewasprotectedfromExoIII digestionafter incu-bation with nuclearextracts (Fig. 5). AP-2plays an

impor-tantrole in neural crest development and as such may be important with respect to transcriptional activity of the XbaI-SphI fragmentinneuronal cells(16). Thepresenceof

anOct-like element in theXbaI-SphI fragmentisinteresting

since neuron-specific Oct proteins that bind HSV-1 IE promotersandinhibittranscriptionhave beenidentified(13). These neuronal Oct-2 factors also differ in their binding specificity and function (6), suggesting thatthis elementor

other similar factorscan influence the activityof the

XbaI-SphI fragment. The HSV-1 latency-associated transcript promoter has a single motif which is a strong enhancer in neuronal cells (3). An 11-base sequence in the XbaI-SphI fragment(CAGGGGCAagG;nt567to557)is theonly region inthe LR promoter which resembles the HSV-1 sequence. SincetheXbaI-AluI fragmentdid not havestrongenhancer activity, this sequence alone doesnot appear tobecrucial for transcriptional enhancer activity. Considering the

dra-maticdifference in enhanceractivityin thetwocelllines,the dataimpliedthat factors inPC-12cells bind sequences in the

XbaI-SphI fragment and activatetranscription. In contrast, nuclear factors in Rat-2 cells bind similar DNA sequences butwereunabletoactivatetranscription.Current studiesare aimed at identifying which factors interact with the XbaI-SphI fragment and how these interactions affect

transcrip-tionalactivity.

ACKNOWLEDGMENTS

We thank D.HamernikandF.Osorio forhelpful discussions and forcriticallyreading the manuscript.

This work was supported by Public Health Service grant R29CA47872 andby USDAgrant92-34103-7168.

REFERENCES

1. Ackerman, M., E. Peterhans, and R. Wyler. 1982. DNA of bovineherpesvirus type 1 in the trigeminalgangliaoflatently infected calves.Am.J. Vet. Res. 43:36-40.

2. Angel, P., M. Imigawa, R.Chiu,B. Stein, R.J.Imbra, H.J. Rhamsdoerf, C. Jonat,P.Herrlich,andM. Karin.1987. Phorbol

esterinducible genescontainacommoncis elementrecognized byaTPAmodulatedtrans-acting factor.Cell49:729-739. 3. Batchelor,A.H.,andP.O'Hare. 1992.Localizationofcis-acting

sequence requirements in the promoter of the latency-associ-ated transcript of herpes simplex virus type 1 required for cell-type-specific activity.J.Virol. 66:3573-3582.

4. Bradford, M. M. 1976. A rapid andsensitive method for the quantitation of microgram quantities of protein utilizing the principle ofprotein-dye binding.Anal. Biochem.72:248-254.

on November 9, 2019 by guest

http://jvi.asm.org/

[image:7.612.109.502.69.296.2]

6106 BRATANICH AND JONES

5. Crosby, S. D., J. J. Puetz, K. S.Simburger, T. J. Fahrner, and J. Milbrandt. 1991.The early response geneNGFI-C encodes a zinc finger transcriptional activator and is a member of the GCGGGGGCG (GSG) element-binding protein family. Mol. Cell Biol. 11:3835-3841.

6. Dent, C. L., K. A. Lillycrop, J. K. Estridge, N. S. B. Thomas, and D.S.Latchman.1991. The B-cell andneuronal forms of the octamer-bindingprotein Oct-2 differinDNA-bindingspecificity and functional activity. Mol. Cell.Biol. 11:3925-3930. 7. Dignam, J. D., R. M. Lebovitz, and R. G. Roeder. 1983.

Accurate transcription initiation by RNApolymerase II in a solubleextractfromisolated mammalian nuclei. Nucleic Acids Res. 11:1475-1489.

8. Falkner, F. G., R. Moickat, and H.G.Zachau.1986. Sequences closely related to an immunoglobulin promoter/enhancer ele-ment occur upstreamof other eukaryotic and of prokaryotic genes.Nucleic AcidsRes.13:7847-7863.

9. Greene, L. A., and A. S. Tischler. 1976. Establishment of a

nonadrenergic clonal line of rat adrenal pheochromocytoma cells whichrespondto nervegrowthfactor. Proc. Natl. Acad. Sci.USA73:2424-2468.

10. Homan, E. J., and B. C. Easterday. 1983. Experimental latent and recrudescentbovineherpesvirus1infectionsincalves.Am. J. Vet. Res. 44:309-313.

11. Jones, C. 1989. The minimaltransforming fragment of HSV-2 is acomplex transcriptionalpromoterelement. Virology 169:346-353.

11a.Jones,C.Unpublished data.

12. Jones, C., G. Delhon, A. Bratanich, G. Kutish,and D. Rock. 1990. Analysis of the transcriptionalpromoterwhich regulates the latency-related transcriptof bovineherpesvirus 1. J. Virol. 64:1164-1170.

13. Kemp, L. M., C. L. Dent, and D. S. Latchman.1990. Transcrip-tionalrepression of herpes simplex virus immediate earlygenes andofoctamercontaining cellularpromotersin neuronal cells is mediatedbythe octamer DNAelement. Neuron4:215-222. 14. Kutish, G., T.Mainprize,and D.Rock. 1990. Characterization

of the latency-related transcriptionally active region of the bovineherpesvirus1 genome. J. Virol. 64:5730-5737.

15. Lo, D. C.1992. NGF takesshape.Curr. Biol. 2:67-69. 16. Mitchell, P. J., P. M. Timmons, J. M. Hebert, P. W. J. Rigby,

and R.Tjian. 1991.Transcription factorAP-2 is expressedin neuralcrest celllineages duringmouse embryogenesis. Genes Dev.5:105-119.

17. Montminy, M. R., K. A. Sevarino, J. A. Wagner, G. Mandel, and R. H.Goodman.1986.Identificationof a cAMPresponsive element within theratsomatostatingene.Proc.Natl.Acad.Sci. USA 83:6682-6686.

18. Rhim,J. S.,J.Fujita,P. Arnstein, and S. A. Aaronson. 1986.

Neoplastic conversion of humankeratinocytes by adenovirus 12-SV40 virus and chemical carcinogens. Science 232:385-388. 19. Rock, D., J. Lokensgard, T. Lewis, and G. Kutish. 1992. Characterization of dexamethasone-induced reactivation of la-tentbovineherpesvirus1. J.Virol. 66:2484-2490.

20. Rock, D. L., S.L. Beam, and J. E. Mayfield. 1987. Mapping bovine herpesvirus type 1 latency-related RNA in trigeminal ganglia of latently infectedrabbits. J. Virol. 61:3827-3831. 21. Rock, D. L., W. Hagemoser, F.Osorio,and D. E.Reed. 1986.

Detection of bovine herpesvirus type 1 RNA in trigeminal ganglia oflatently infected rabbits by in situ hybridization. J. Gen. Virol. 67:2515-2520.

22. Rock, D. L., and D. E. Reed. 1982. Persistent infection with bovineherpesvirus1:rabbitmodel. Infect. Immun. 35:371-373. 23. Ronnett,G. V., L. D. Hester,J. S. Nye, K.Connors,andS. H. Snyder.1990. Humancortical neuronal cell line: establishment fromapatient with unilateral megalencephaly. Science 248:603-605.

24. Salton, S. R. J., D. J. Fischberg, and K.-W. Dong. 1991. Structureof thegeneencoding VGF,anervoussystem-specific mRNAthat israpidly andselectively induced bynervegrowth factor in PC12cells. Mol. Cell. Biol. 11:2335-2349.

25. Sambrook, J., E. F. Fritsch, and T.Maniatis.1989. Molecular cloning:alaboratory manual,p.1:21-1:50.ColdSpring Harbor Laboratory,Cold Spring Harbor,N.Y.

26. Stein, B., H. J. Rahmsdorf, A. Steffen, M. Litfin, and P. Herrlich. 1989. UV-induced damage isanintermediate stepin UV-inducedexpressionof humanimmunodeficiencyvirustype 1, collagenase, c-fos, and metallothionein. Mol. Cell. Biol. 9:5169-5181.

27. Williams,T., and R.Tjian.1991.Analysis of theDNAbinding and activation properties of the human transcription factor AP-2.GenesDev.5:670-682.

28. Wirth, U. V.,C. Fraefel,B.Vogt,t.Vlcek, V. Paces, and M. Schwyzer. 1992.Immediate-earlyRNA 2.9andearlyRNA2.6 of bovine herpesvirus 1 are 3' coterminal and encode aputative zincfinger transactivator protein.J.Virol. 66:2763-2772. 29. Wirth, U. V., B. Vogt, and M. Schwyzer.1991.The three major

immediate-early transcripts of bovine herpesvirus1arisefrom

twodivergent and spliced transcription units. J.Virol. 65:195-205.

30. Wu, C. 1985. An exonuclease protection assay reveals heat shock element andTATAboxbinding proteinsincrude nuclear

extracts.Nature(London)317:84-87.

31. Wyler, R., M. Engels, and M.Schwyzer. 1989.Infectious bovine rhinotracheitis/vulvovaginitis (BHV-1),p.1-72.InG.Wittman andY. Becker(ed.), Herpesvirus diseases of cattle, horses, and pigs. Developments in veterinaryvirology ser. Kluwer Aca-demicPublishers,Boston.

J. VIROL.

on November 9, 2019 by guest

http://jvi.asm.org/