Structural and functional characterization of rev-like transcripts of equine infectious anemia virus.

Structural and

Functional

Characterization of rev-Like

Transcripts of

Equine Infectious Anemia Virus

RINA

ROSIN-ARBESFELD,l

MICHAL

RIVLIN,l

SILVIA

NOIMAN,1

PNINA

MASHIAH,'

ABRAHAMYANIV,' TORU MIKI,2 STEVEN R. TRONICK,2* ANDARNONA GAZIT' Department of Human Microbiology, Sackler School of Medicine, Tel Aviv University,

Tel Aviv 69978, Israel,1andLaboratory ofCellular and Molecular Biology, Building 37, Room 1E24, National Cancer Institute,

Bethesda, Maryland 208922

Received 1 March 1993/Accepted 18 May 1993

Three cDNA clones representing

structurally

distinct transcripts were isolated from a cDNA library

preparedfrom cells infected withequineinfectiousanemia virus(EIAV) byusingaprobe representing the S3 openreading frame,which isthoughttoencode Rev.Onespecies, designated p2/2, contained fourexonsand wasidenticaltoa

previously

described polycistronicmRNAthat encodes Tat. This transcriptwaspredictedto alsodirect the synthesisofatruncated form ofthetransmembrane protein andaputative Rev protein whose N-terminal 29 amino acids, derived from env, are linked to S3 sequences. The second cDNA, pl76, also consisted of fourexonswhichweregenerated bytwoof three of thesamesplicingeventsthatoccurwithp2/2 but notwith the Tat mRNA. The alternative splice sitegiving rise tothe second exon ofp176 results in a

bicistronicmessagethat would encode thesametransmembraneand Revproteinsasp2/2.The firstexonofthe thirdtranscript, p20,wasidenticaltothose ofp2/2andp176butwasspliced

directly

toS3.This monocistronic messagecould encodeasecondform ofRev that lacksenvsequences,providedthatRevsynthesiswould initiate atanon-AUG codon.Thecoding capacityofeachcDNAwasassessed inaeukaryoticsystemusingS3antisera. TwoputativeRevproteinswith apparentmolecularmassesof 18 and 16kDawereexpressedby p2/2andp176,

while p20 expressed only a 16-kDa species. Analysis of EIAV-infected cells with S3 antisera revealed the presence ofan 18-kDa protein.

Surprisingly,

the sameproteinwas detected in purified virions. By usinga reporterconstruct,thechloramphenicolacetyltransferasegenelinkedtoEIAVenvsequences,we wereableto demonstrate greatly enhanced chloramphenicol

acetyltransferase

activity

in cells cotransfected with this constructandanyof the threecDNAs.

Thegenomeofequineinfectious anemia virus(EIAV)(20, 42),alentivirus,isstructurallylesscomplexthan those of all other members ofthis virussubfamily (14).Inadditiontothe threemajoropenreadingframes(ORFs),gag,pol,andenv, threesmallORFs, designated Si, S2,andS3,arepresent.Si encodessequencesrequiredfor the function of the EIAVtat

gene (11, 33, 45). There ispersuasive but indirect evidence thatS3 sequencesrepresentrev(35, 47),whereas the func-tion ofS2 is unknown. The pattern of EIAVgeneexpression

in infected cells belies its simple genomic structure and resembles that of its muchmore complexcounterparts (8).

Thus, at leastfivespeciesofvirus-specific transcriptswere observed (34), and analysis of cDNA clones has revealed that the smaller bands present in Northern (RNA) blots

actually represent multiple mRNA species (33, 35). In our

analysis of cDNAlibraries ofan EIAV-infected canine cell line (33, 35),wedemonstrated that theEIAV Tatproteinis encoded by at least three alternatively spliced transcripts.

One mRNAwasshown tobepolycistronic, encodingTat, a

putative Rev, and/or a truncated transmembrane (ATM)

protein. Another bicistronic tat message that could also potentially direct the synthesis of a truncated TM protein was isolated. A monocistronic tat mRNA was also de-scribed. We found that the relative abundance of these transcripts differed in infected cells, and trans-activation assaysshowedthat thetatactivityof themonocistronicform

*Correspondingauthor.

was significantly higher than that of the more complex

species (35).

Here we present the characterization of putative EIAV Revtranscripts isolated from cDNA librariespreparedfrom EIAV-infected cells by screening for recombinant phages containing S3 sequences. Although one clone, designated

p2/2, was found to be identical to the previously reported polycistronictat clonep105 (33), two other distinct cDNA

clones, designated p176andp20,werefound. These cDNAs were isolated by using pCEV vectors which allow highly

efficient directionalcloningof functional full-length cDNAs whicharereadilyconvertedintoaplasmidform forrescuein bacterial cells(30).Forsequenceanalysis,the cDNA inserts were subcloned into pBluescript vectors (Stratagene) and

analyzedwith the Sequenase DNA sequencingkit (United

States Biochemical Corp.), as previously described (35). Clonep176 (Fig. 1)was1,361 bplong,and its first nucleotide

correspondedtoposition223 of the EIAVgenomicsequence (20),which is 16 bpdownstream of the transcription start site. Comparison of the nucleotide sequenceof this cDNA with thegenomicsequencerevealed that itwascomposedof fourexons(Fig.1 and2).Exon 1 extendedtotheconsensus

splicedonor siteatposition459. The secondexonstartedat base 5210, located approximately in the middle ofS1 and upstream of env, and terminated at the splice donor site

previouslyshowntobe used togeneratetatmRNAs

(posi-tion5276) (33, 35). The thirdexonstarted within S2atbase 5437 and extendedtobase 5534. The fourthexonstartedat

the firstbase of S3(position 7232) and endedatthepoly(A)

site within the long terminal repeat. Thus, this cDNAwas

5640

on November 9, 2019 by guest

http://jvi.asm.org/

10 20 30 40 50 60 70 80 90 100 CGGTCTGAGTCCCTTCTCTGCTGGGCTGAAAAGGCCTTTTAATAAATATAATTCTCTACTCAGTCCCTGTCTCTAGTTTGTCTGTTCGAGATCCTACAG

110 120 130 140 150 160 170 ISO 190 200

TTGGCGCCCGAACAGGGACCTGAGAGGGGCGAGACTACCTGTTGAACCTGGCTGATCGTAGGATCCCCGGGACAGCAGAGGAGAACTTACAGAAGTCTTC

Exon 1 4591 4S Exon 2 [52101

210 220 230 240 250 260 270 280 290 300

TGGAGGTGTTCCTGGCCAGAACACAGGAGGACAGAGGAAGAAGAATA^A AAAGACTGAAGGCAATCCAACAA(GGAAGACAACC-TC.AATA TT TGTTAIAA

II K11 e G N

Tm

P T

I1)VH~

R K S T

Exon 2

[5276

Sl)

T (altrnate frame in Exon 4

(72321

Exon 3 [5

41

7 Exon 3

[55341j

'

4-

-310 320 330 340 350 360 370 380 390 400

t

(SU

codon 43)

'.'.'.''

.'''...'

K.'

''.K

'.'.-.K-.

K'''"''

K-'V'

K ''I''''.'#.'-'"

t (alternate frame in SU)

410 420 430 440 450 460 470 480 490 500

D P Q G P LWSE S W C R V P E E K I P S Q T C I A R

...* .. ,.

t (TM codon 197)

S10 520 530 540 550 560 570 580 590

600

IHF L A P G. P T 0 T P s R R D R W T R a 0 T L a T E V L F R

LF.I

...-- ---WK

610 620 630 640 650 660 670 680 690

700

rAr,AATrR-ArAr. r. ,TArlAAAeRlRlRRrIMrs AAAr,Ar,r.r.r^RTT Arer r.T* Arr,CAT TTR rARAR^r,Ar, TT T^n^R

ARnn

*,

*RAr T

I RGB V QQAAKFn K rFA P E L Y F R E R G D I9

710 720 730 740 750 760 770 780 790 800

~ .6.

A*F

p AL

-P

R V L s

810 820 830 840 850 860 870 880 890 900

r,-TTT,*TGACTGTTGCATTAAAGCCCAAGAAGGAACTCTCGCTATCCCTTGCTGTGGATTTCCCTTATGGCTATTTTGGGGACTAGTAATTATAGTAGGA

...--..W...

910 920 930 940 950 960 970 980 990 1000

CGCATAGCAGGCTATGGATTACGTGGACTCGCTGTTATAATAAGGATTTGTATTAGAGGCTTAAATTTGATATTTGAAATAATCAGAAAAATGCTTGATT

5.14.0V"~~~~'L'* * 4. A ~~~~ V ~~1 ft '1 0 1 ft S N.. LI. ..I....I .~~..-...~~~~~~~.I...5 1 .1.

1010 1020 1030 1040 1050 1060 1070 1080 1090

1100

ATATTGGAAGAGCTTTAAATCCTCGCACATCTCATGTATCAATGCCTCAGTATGTTTAGAAAAACAAGGGGGGAACTGTGGGGTTTTTATGAGGGGTTTT

lio 1120 1130 1140 1150 1160 1170 1180 1190

1200

ATAAATGATTATAAGAGTAAAAAGAAAGTTGCTGATGCTCTCATAACCTTGTATAACCCAAAGGACTAGCTCATGTTGCTAGGCAACTAAACCGCAATAA

1210 1220 1230 1240 1250 1260 1270 1280 1290 1300

CCACATTTGTGACGCGAGTTCCGCATTTGTGACGCGTTAAGTTCCTGTTTTTACAGTATATAAGTGCTTGTATTCTGACAATTGGCACTCAGATTCTGCG

1310 1320 1330 1340 1350 1360

GTCTGAGTCCCTTCTCTGCTGGGCTGAAAAGGCCTTTGTAATAAATATAATCCTCTACTCA

FIG. 1. Nucleotidesequence and deduced amino acidsequenceofcDNA clonep176.Nucleotide positions of the cDNA are shown above thesequence, and thecorrespondingpositions intheEIAVgenomic sequence, indicating exon borders (20), are bracketed. The REV ORF

(nt249to806) (openbox), sequences derived fromalternatephases of Si (nt 249 to 300), SU (nt 301 to 400), and S3 (nt 401 to 806), a potential initiationcodon (nt 312) (boxed), theATMORF(nt355to1056)(shaded), and sequences derived fromalternatephases in SU (nt 355 to 399) and TM(nt 400 to 1056) areindicated. The figure waspreparedwith the aid of the computer program DNAdraw, written by M. Shapiro, National Institutesof Health.

similarto

p2/2

but could not encode a Tat proteinbecause of

theuseofa

splice

site well within S1 (Fig. 2). The first exon

of

p20

wasidentical to that ofp176 but was spliced directly to the second codon of S3 (position 7235) (Fig. 2).

The sequence ofp2/2 (data not shown)is identical to that

of clone

p105

(33, 35),

which encodes Tat and is predictedto

express Revand atruncated TM protein. Clonep176 (Fig. 1

and

2)

contained

only

the ORFs REV and ATM. In theREV

ORF

(nucleotides

[nt]

249 to 806), there is an AUG codon

(residue

22)

in a favorable context (22, 23) for an initiation codon. If

translation

initiates here, the product would be

predicted

toconsist of 165 residues (19,760 Da),with its first

29

amino

acids derived from env and fused to 136 codons encoded

by

S3. The second ORF, ATM (nt 355 to 1056),

containsa

potential

AUG initiation codon at position11, but

its context is suboptimal (23). If this initiation codon is

functional,

a

protein

of 224 amino acids (24,995 Da) would result. Its N-terminal five amino acids are derived from an

alternate

phase

of the envsequence and are linked directlyto

codon 197 of the TM ORF. Rice et al. (40) found that the EIAV TM

protein

is cleaved at residue 240 into N-terminal

glycosylated

and C-terminal

nonglycosylated

products

des-ignated

gp32

and

p20,

respectively. These proteins were

EIAVGENOME

LTR ga

Upo pot

S2 LTR

U

p2/2

-evRevA

Tat

---;E1---p

Rv2p20

FIG. 2. Structuresof putative Rev cDNAs. The organization of theEIAVgenome, the surface glycoprotein (SU) and TM coding regionswithinenv, and the exons (black boxes) and predictedORFs

(patterned boxes) of the Rev cDNAs are illustrated. Patterns on boxescorrespondtothepatterns used for the EIAV genomic coding regionsatthetop.The white boxes that start the ATM ORF indicate analternate phase inenv. LTR, long terminal repeat.

F37277777,7777773..,/,

Si EE]

on November 9, 2019 by guest

http://jvi.asm.org/

[image:2.612.68.563.98.412.2] [image:2.612.324.568.511.661.2]

10 20 30 40 50 60 70 80 90 100

110

120

130

140

150

160

170

180

190

200

CGCCCGAACAGGOACGAGGQAATACCTGATTGAACCTGGCTGATCGTAGGA.LCCCCGGGA.CAGCAGAGGAG

AACTTACAGAAGTCT TCTGGA

I0PR

0

S

R

G

E L T

EVFW1

21

1

Ext

2

30

E 2o242250

260

2

70

2

80

2 90

300

GGTGT

TCCTGGC,CA.G.AA

A AG

A.&A,,A.T

TQTAGCTTG

QA.C.C.A.G.TGGT.GCA.G.G.T.CC.T.C.C.G.GCAG.TC.GT.T.A.C.CTGAA.G.A.A.A.AAAT

t (TU flln lAfl

t (S3 codonf 1)

3 10

32

3

30

3

40

3

50

3

60

3 70

3

80

3 90

40 0

410 4

20

4

30

4

40

450

4860

4

70

4

80

4

90

50 0

[image:3.612.56.551.73.361.2]

ACAGAAGTAC

TCCAG.G.A

C

ATGGikA.TGGAG.AAT.CAGA

GGAGTACAACA

GGCG%GC.CA

A.AGA

GCT

GGGTGA.AGT.C.A.A.T.C.G.AGGC.A.T.T.T.GG.A.G.AG.A.GC.T.AT

... ...44 t t t t W V X ...A

510

529

530

540

550

560

570

580

590

600

A....C-.A...CCAA Q..

TTT.ACTAG0.G...CAACOACACAAGA....CTCT...06..GA.c....CCT.ACCAAGGGTCCT..

7A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...

FRA-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... ...

610 egb. 630 640 650 660 670 680 690 700

AC.C.TQGA.GA.T.T. TTTATOIIAC

CO A.

A,2l..QJQ...&&Qj.AGRA.A.

6.A.C.A ... .... .... lix,

S. F.

P G D S K R R R K H L

710 72

0

730 740 750 760 770 780 790 800

A.T.TT.T.G.GQG.A.C.T.A.G.TA.A,T.T-4.1.A.G.T.A.G.GAC.G.CA.T.A..G.C..A.G.GC.T.A.TGG.A.T.T.A.C.G.T.O.G.A.C.T.C.G.C.T.G.TT.A.T.AA.T.A.AG.G.A.T..T.T.G.TAT.T.AGA.G.GCT.T.AA.A.T.TT.G.A.TA...V... ..

... ... ...a... ...L... ... ... .. ... ... ..._v... . V A %: %

810 820 830 840 850 860 870 880 890 900

T.T.TQA.A.A.T.A.A.TCA.G.AA.A.A,... X A-TOCTT.G.A.TT.A.T.AT.T.G.GA.A.G.A.GC.T.TT.A.A.A.T.C.C.TGG.C.A.CA.T.C.T.C.A.T.G..T.A.T.C.A.A.T.GC.C.T..C.A.G.TA..T.G.T.T.TAGAAAAACAAGGGGG... .3L ...:x, ..T. -Xm... y

V--910

920

930

940

950

960 970

980

990

1000

1010

1Q20-

1030

1040

1050

1060

1070

1080 1090 1100

1110

1.12

1130

1140

1150

1160

1170

1180

TTCTGACAATTGGGCACTCAGATTCTGCGGTCTGAGTCCCTCTCTGCTGGGCTGAAAAGGCCTTTGTAATAAATATAATTCTCTACTCA

FIG. 3. Nucleotide

sequence

and deduced amino acidsequence of cDNA clone

p20.

The REV ORF

(nt

158to

634) (open

box),

sequences derivedfromexon1

(nt

158to

230)

and S3

(nt

231to

634),

the ATM ORF

(nt

120to

884) (shaded),

sequences derived fromexon1

(nt

120to

230)

and from TM

(nt

231 to

884),

and the codon used toinitiate the Tat

protein

(nt

159)

(boxed)

areindicated.

found in virions.

Whethier

the

putative

truncated

envelope

proteins

encoded

by

the. EIAV cDNAs described here have afunction

analogous

tp

thiat

of

p20

remainstobedetermined. A truncated

envelope

protein

was also

suggested

to be encoded

by

therev

transcripts

of visnavirus

(29)

and

caprine

arthritis

encephalitis

viru's

(19).

Itis ofinterest that C-termi-nal sequences

present -in

EIAV and other lentiviral TM

proteins

possess

structuires

similar to those of a

family

of

cytolytic

peptides

andcould

play

arole in

cytopathic

effects induced

by

lentiviral

-infection

(31).

Clone

p20

is

predicted

tocontaintwoORFs

(Fig.

2and

3).

The first

ORF,

designgated

ATM

(nt

120 to

884),

lacks an obvious translation initiation

codon,

but it is

intriguing

that its first 38 codons are identical to those of the tat gene

product

andare

spliced,

tocodon 198 of the TM ORF.

Thus,

atruncated TM

protein

of

26,576

Da could be

synthesized

by

using

the Tat

protein's

initiation

codon

(isoleucine,

AUC)

at

position

14

(35). Interestingly,

Beisel et al.

(3)

detected a

transcript analogous

to

c~DNA

p20

in horses

acutely

infected with EIAV. The second

ORF,

designated

REV,

(nt

158 to

634),

starts 66

bp

downStream

from the 3' end of the

long

terminal

repeat.

This

stre'tch,

whichlacks anAUG

codon,

is

spliced

tothe second co'don of

S3.

Thus,

the

product

ofthis ORF would

potentially,represent

a form of Rev with an amino terminusdifferent

from

that encoded

by p176

(assum-ing

that initiationoccurs at anon-AUG

codon).

The

coding capacity

of~

each cDNAwas assessed

by

the use of a

coupled

in vitro

transcription-translation

system.

Each

cDNA,

subclorned

into a

Bluescript

vector,was

tran-scribed in vitro with a

Stratagene

kit,

and the

resulting

capped

RNAswere usedtoprogram a

reticulocyte

transla-tionextractinthe presence of

[35S]methionine

and

[35S]cys-teine

(35).

Two

proteins

with

apparent

molecularmassesof 20 and 17 kDawere

synthesized

in response to clone

p2/2

DNA

(Fig.

4A,

lane

a).

The differences in the sizes of the bands

(20

and 17 kDa in vitroversus 18 and16 kDa in

vivo)

are

likely

dueto

gel

and

sample

composition variability.

The 8-kDa Tat

protein

(35)

was not

observed,

presumably

be-cause ofthe inefficient use of non-AUG codons in in vitro translation

systems

(23).

The sameresultwasobtained with clone

p176

(Fig.

4A,

lane

b).

No

cDNA-specific

bandswere observedinextracts

programmed by

clone

p20

(Fig.

4A,

lane

c).

This result may also be

explained

by

thelack ofanAUG codon in the

predicted

ORFs of

p20.

To further characterize the

products expressed by

clone

p176,

a

pool

of anti-S3sera

(antiserum

177,

whichwasraised

against oligopeptide

177

[LRQSLPEEKIPSQTY],

and anti-serum

181,

raised

against oligopeptide

181

[DFSAWGGY

QRAQERL])

wasused.Thesametwo

proteins

were

specif-ically immunoprecipitated by

anti-S3 sera

(Fig.

4A,

lane

e),

suggesting

that both

proteins

weretranslated from the REV ORF. Their sizes were most consistent with initiation at

positions

22and 33 of the REV ORF.

Among

other

possi-bilities,

both

might

start atcodon 22 and the differences in

apparent

molecular

weights

might

be due either to RNA

editing

(18)

ortodifferentialmodification

by

phosphorylation

(9)

Next,

we

investigated

the

coding capacities

of the three cDNA

species

in a

eukaryotic expression

system.

The cDNAs were cloned into amammalian

expression

vector,

pMAMneo

(Clontech),

downstream of the dexamethasone-inducible

long

terminal

repeat

of mammary mouse tumor

on November 9, 2019 by guest

http://jvi.asm.org/

B

a

b

e

"i

..

...:.. ..:. ...

d

<201

4 17

a

b

c

d

a

18

4 16

C

9

a

b

c

kmrn

_

Ial

FIG. 4. Coding capacity of the Rev transcripts. (A) Analysis of in vitro translation products of cDNAs. [35S]methionine- and [35S]cysteine-labelledpolypeptidesweresynthesizedinanin vitrotranscription-translation reaction mixturecontaining5 ,gof DNA from pKS176 (lane a), pKS2/2 (lane b),orpKS20 (lane c)orfrompKS2/2in theantisenseorientation(lane d),andtheproductswereseparated bysodiumdodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (15%

[wt/vol]

polyacrylamide gel). In asecondgel, products of the reactionprogrammed by pKS176DNAwerepreadsorbed with preimmuneserum(1hat0°C)and thentreated withprotein A-Sepharose(30 minat0WC). Supernatantswerethenimmunoprecipitatedwithapoolofanti-S3sera(no.177and181) (lane e)oranti-S3serapreadsorbedwith theappropriate oligopeptides (lane f).Arrowsindicate thepositions ofthe17- and 20-kDaproducts. (B) Analysis of cDNA expressioninCOS cells. COScellsweretransfected withpMAM/176 (lane a), pMAM/2/2 (lane b), pMAM/2O (lane c),orpMAM (lane d). Cellsweregrownfor 32daysin thepresenceofG418 (900pg/ml) and then treated for48 h withdexamethasone(10-6 M). Lysatesfrom 6x106 cellswereincubated

first with preimmune serum (1 h, 0°C) and then with protein A-Sepharose (30 min, 0°C) before immunoprecipitation of the resultant

supernatantwithapoolofanti-S3sera(no. 177 and 181). Inasecond gel, celllysatesofpMAM/2/2-transfectedcellsincubatedinthepresence (laneseandg)orabsence(lane f)ofdexamethasonewereadsorbed withpreimmuneserum asdescribed aboveand thenimmunoprecipitated with anti-S3sera(laneseand f)oranti-S3serapreadsorbed with the appropriate oligopeptides(lane g). Proteinsweredetectedby using the poolofanti-S3seraand

"2I-labelled

proteinA.Arrowheadsindicate thepositionsof the16- and 18-kDaS3-encodedproducts. (C)Analysis

of EIAV-infected cells and virions. Totalcelllysatesprepared from5 x 106persistentlyinfected(lane a)oruninfected (lane b)canine cells andSDS-lysed, gradient-purified EIAVvirions(lane c)were treated with preimmune serum,immunoprecipitatedwith anti-S3 sera,and immunoblotted asdescribedabove.

virus. The three cDNA expression vectors pMAM/2/2, pMAM/176, andpMAM/20weretransfected into COScells.

Following 32 days ofgrowth in the presenceof G418 (900 ,ug/ml),cellsweretreated withdexamethasone(10-6M) and

48 hlater, cell extractswere immunoprecipitated and then

probedwith apoolof thetwo S3antisera. Twoproteins of approximately16 and 18kDawereimmunoprecipitated from cells transfected with pMAM/2/2 or pMAM/176 (Fig. 4B, lanes a and b), while only one protein of 16 kDa was

immunoprecipitated from cells transfected with pMAM/20 (lane c). In contrast,noequivalentbandwasobserved with cells transfected with the vectoritself, nor was there any S3-specific signal when the S3 antisera were preadsorbed

with the immunizing peptides. The S3 antisera were next usedtoanalyzeS3-encodedproteinsinEIAV-infected cells.

Figure 4C (lane a) shows that an 18-kDa protein, which comigratedwith the S3-encoded 18-kDaprotein of pMAM/ 2/2 (data not shown),was specifically immunoprecipitated

from extracts of infected cells. Stephens et al. (47) previ-ously reported an S3-encoded 18-kDa protein in EIAV-infectedequine fetalkidney cells.

Since itwasreportedthat the Revproteinof visnavirus is presentinmaturevirions(29),itwasofinteresttodetermine

whetherthiswasalsotruefor EIAV.Supernatantsof canine

cellspersistentlyinfected with EIAVwerecollectedat24-h

intervals.Virionswerepelleted,and theviralproteinswere

immunoprecipitated bythepoolof S3antisera. Aproteinof 18 kDawas specificallydetected in EIAV virions (Fig. 4C,

lanec).The16-kDa band present(Fig. 4C,lanec)islikelyan

artifact,sinceitwasnotdetected whenvirionextractswere first immunoprecipitated and then analyzed by Western

immunoblottingwith the anti-S3 sera. Furthermore, afaint

16-kDabandcanbeseenin thecontrol lanesontheoriginal autoradiogram.

The results of the present study indicate that the three

speciesofcDNAs, p2/2, p176, andp20, expressed proteins

encodedbyS3. Theyallgaverisetoacommonproteinof16 to17kDa, and two,p2/2 andp176, expressedanadditional protein of 18 to 20 kDa. It is noteworthy that the Rev

transcriptof visnavirus alsoencodestwospeciesof Rev-like

proteins (29). On the basis ofsequenceanalysisof the three

transcriptsandtheirpredicted coding capacity,it ispossible

thatthe 18-kDa S3-encoded protein expressed by p2/2 and

p176 is a protein of 165 amino acids whose 29 N-terminal residues are derived from the surface glycoprotein (SU)

sequences(codons43to73)whicharefused in frameto S3.

The structure of this protein is analogous to those of the

predictedRevproteinsof otherungulate lentivirusessuchas visna virus (13, 29, 44), visna-like ovine virus SA-OMVV

(39), bovine immunodeficiency virus (12, 36), and caprine

arthritis encephalitisvirus(19, 43).

Alignmentofthepredicted18-kDaS3-encodedproteinof EIAV with the Rev proteins of other lentiviruses (47) re-vealed the presence of three conserved domains. The first domain contains a cluster of charged amino acids rich in

glutamylresidues(KEESKEEKRRNDWWKK, residues36 to 51 of the REV ORF). Although the N termini of the

primateRevproteinsarenotderived fromenv(6, 7, 16), they contain an analogous cluster ofcharged residues (RSGDS DEDLLKAVRLIK). Recent mutational analysis (17) sug-gestedthat this N-terminal stretchis involved in thebinding specificityofthe Revprotein for the Revresponseelement (RRE). Thesecond domain is also a highly charged region

composedofseveral basicamino acids in the secondcoding

A

on November 9, 2019 by guest

http://jvi.asm.org/

[image:4.612.88.548.71.228.2]

A

B

a b c a b c d

0

C

0 30 90 150 210

TIME(min)

FIG. 5. Assay forRevactivity. The CAT reporterplasmid pCMV-CAT-RREwasconstructedby cloning the CAT coding region (the BglII-to-Bsp 1286 Ifragment whichspansnt-42to689, where+1referstotheAof the CAT initiationcodon) into theBamHI-HindIIIsites downstreamof theCMV promoter inpCEV21(36). Aportion of theenvregion of theEIAV provirus(theBglII-to-ApaI fragment [nt 4473 to7247]), which presumably contains theRRE but lacks thepresumedRevcodingsequence,wascloned into theBglII-ApaIsitesofpCEV21, downstream of the CATregion. Theinsulinpoly(A) signalis located downstream of the insertedenvregion.Asimilar construct,designated pCMV-CAT,whichlacksthe envregion, servedasacontrol.(A) Canine cellsweremock transfected(lane a)ortransfected with4 ,ugof DNA frompCMV-CAT-RRE(laneb)orpCMV-CAT(lanec). After48h, CAT levelswereassessedaspreviously described (45).Percentagesof

[14C]chloramphenicol

convertedto acetylatedderivativeswere asfollows: lane a,0; laneb, 23.0; and lane c, 98.1.(B) Caninecellswere

co-transfectedwith 4 ,ugofDNAfrom pCMV-CAT-RREand 4 Lg of DNA from cDNA clone

p20

(lane a), p2/2 (lane b),or

pl76

(lane c)or

the pCEV21vector (laned). After48h,CATlevelswereassessedand the percentages of[ C]chloramphenicol converted toacetylated derivativeswere asfollows:lane a,89.3; laneb, 98.4;lane c,99.2; and lane d, 35.9. (C) Duplicate samplesweretransfectedasdescribed for panelsAand B,andCATactivitywasdeterminedby themethod of Neumannetal.(32). Symbols: U,clonep176;El,clonep2/2;*,clone p20; 0,pCEV21.

exon

(RRDRWIRGQ, residues 97

to 105 of the REV

ORF),

reminiscent of the arginine-rich

motif

of the Rev

proteins

of primate lentiviruses. These residues are

thought

to be in-volved in binding to the RRE

(21,

25,

27)

and are also

thought

to act asthenuclear localization

signal

(4, 26, 38, 51,

53).

The third conserved

domain,

located at the

carboxy

terminus of the

putative

S3-encoded Rev

protein,

spans residues 129 to 141

(ELGEVNRGIWREL)

and may be analogous tothe human

immunodeficiency

virus type1 Rev

sequence

(LPPLERLTLDCNED),

whichwas

suggested

to

functionas anactivation domain that bindstocellular factors involved in RNA transport

(24, 26, 37, 50).

The 16-kDa S3-encoded

protein,

which is

expressed

by

all three

transcripts, presumably

initiatesat anon-AUG codon locatedatthe 5' end of S3.Consequently,

this

protein would lack the

highly charged

N-terminal

glutamic

acid-rich do-main locatedatthe5' end of the 18-kDa

protein.

However, it wouldcontain the othertwoconserved

regions

encodedby S3.

To further substantiate thenotion that the three

species

of cDNAs representrev

transcripts,

weexamined whether the

proteins

expressed by thesecDNAscouldovercomeintrans

the effects of env sequences in

inhibiting expression

of a

chloramphenicol acetyltransferase

(CAT)-env

reporter con-struct

(38).

Areporter

plasmid

(pCMV-CAT-RRE)

in which

aportion ofenvlacking S3, but presumablycontaining the RRE, located downstream of the CAT coding sequence and under the control of thecytomegalovirus (CMV) promoter,

was constructed. The CAT activity of this construct was

23%

conversion (versus 97.2% for pCMV-CAT) (Fig.

5A).

Significantly higher

CAT levels wereobserved in the

pres-ence of each one of the rev cDNAs. Clone

p20,

which expresses a 16-kDa S3-encoded

protein,

enhanced CAT production by pCMV-CAT-RRE to a level of

76.8%

conver-sion.

Moreover,

cDNA clones

p2/2

and

p176,

which express

18- and 16-kDa

protein

species, elevated CAT levels to 95.4 and

97.9%, respectively.

Using

theFOLD programas

implemented

inthe

Univer-sity

of Wisconsin Genetic

Computing

Center's sequence

analysis

software

package,

we

performed

exhaustive

analy-ses of the EIAV env nucleotide sequence in attempts to

identify

a

potential

RRE;

however,

none was found.

Thus,

the EIAVRRE

might

havea much more subtle

secondary

structure

than

thoseidentified in other lentiviruses or may

use an

altogether

different sequence motif. In any case, it will be necessaryto

identify

the EIAV RREby recombinant

techniques and/or

direct RNA

binding experiments.

It will also be of interest to determine whether the EIAV Rev

protein

canrecognizethe RREs of other

lentiviral

genomes. Our data indicate that the Rev

proteins

of EIAV are

expressed

via

alternatively spliced

mono-and

polycistronic

messages, asis thecasefor otherprimate

(2,

15, 41, 49,

52)

andungulate

(10,

19,

29, 36, 43)

lentiviruses. However, the

polycistronic

messagesofEIAV may expresstwo

species

of Revwith distinct N termini. In other systems, it has been shown that the use of alternative

initiation

codons from within thesamemessageresults in differential

protein

local-ization

(1,

5,

46).

It remains to bedetermined whether the

two

putative

Rev

proteins

arelocalized

differently.

The Rev

protein

of visna virus

(19 kDa),

in contrast to

primate

lentiviral Rev

proteins,

accumulated

preferentially

innuclei

(48)

aswellasinthe

cytoplasmic

and membrane fractions of infected cells

(28).

Moreover, the largest

species

of the Rev protein of visna virus (19 kDa) was packaged into mature

viral

particles,

as is shown here for EIAV. Thus, the Rev

protein might

playarole in theassembly of viral particlesor

in the

early

stepsofvirus

infection,

in additionto

regulating

the

cytoplasmic

expression of

unspliced

or

singly

spliced

viral

transcripts.

Nucleotide sequence accession numbers. The nucleotide

;VN-,El 2

?n ) # lp

.i-;%', r", E

u

4b 4'o

on November 9, 2019 by guest

http://jvi.asm.org/

[image:5.612.129.478.72.231.2]

sequences

and deduced amino acid sequences of cDNA

clones

p176

and

p20

have beenlodged in the GenBank data

base under accession numbers X63059 and X63058,

respec-tively.

We thankS. A. Aaronson for continued support and

encourage-ment.

This studywas supported by a project grant from the United

States-Israel Binational ScienceFoundation. REFERENCES

1. Acland,P.,M.Dixon, G.Peters, and C. Dickson. 1990. Subcel-lular fate ofthe Int-2 oncoprotein is determined by choice of initiation codon.Nature (London) 343:662-665.

2. Arrigo,S.J.,S.Weitsman, J.A.Zack, and I. S. Y. Chen. 1990. Characterization and expression of novel singly spliced RNA species of human immunodeficiency virus type 1. J. Virol. 64:4585-4588.

3. Beisel, C.E., J.F.Edwards, L. L.Dunn, and N. R. Rice. 1993.

Analysis

ofmultiplemRNAs frompathogenic equineinfectious anemia virus(EIAV)inanacutely infected horsereveals a novel

protein,

Ttm, derived from thecarboxyterminus of the EIAV transmembraneprotein. J.Virol.67:832-842.

4. Berger, S., C. Aepinus, M. Dobrovnik, B. Fleckenstein, J. Hauber, and E. Bohnlein. 1991. Mutational analysis of func-tional domains in the HIV-1 Rev trans-regulatory protein.

Virology

183:630-635.

5. Bugler, B.,F.Amalric,and H. Prats. 1991.Alternativeinitiation oftranslation determinescytoplasmic ornuclear localization of basicfibroblastgrowth factor. Mol. Cell. Biol. 11:573-577. 6. Chakrabarti, L.,M. Guyader, M.Alizon, M. D. Daniel, R. C.

Desrosiers,P.Tiollais, and P.Sonigo. 1987. Sequence of simian immunodeficiencyvirus from macaque and its relationship to otherhuman and simian retroviruses. Nature(London) 328:543-547.

7. Colombini, S., S. K. Arya, M. S. Reitz, L. Jagodzinski, B. Beaver,and F.Wong-Staal. 1989. Structureofsimian

immuno-deficiency

virus regulatorygenes. Proc. Natl. Acad. Sci. USA 86:4813-4817.

8. Cullen, B. R. 1992.Mechanism of action of regulatory proteins encodedby complexretroviruses. Microbiol. Rev. 56:375-394. 9. Curran, T.,M. B.Gordon,K. L.Rubino, and L. C. Sambucetti. 1987.Isolationandcharacterization ofthec-fos (rat) cDNA and

analysis

of post-translational modification in vitro. Oncogene 2:79-84.

10.

Davis,

J. L., andJ. E. Clements. 1989. Characterization of a cDNA clone encoding the visna virus transactivating protein. Proc. Natl.Acad.Sci. USA86:414-418.

11. Dorn, P.,L.DaSilva,L.Martarano, and D. Derse. 1990. Equine infectiousanemia virustat:insights intothe structure, function, and evolution of lentivirus trans-activator proteins. J. Virol. 64:1616-1624.

12. Garvey,K.J.,M.S.Oberste, J.E. Elser, M. J.Braun, and M. Gonda. 1990.Nucleotidesequenceand genome organization of

biologically

activeprovirusesof thebovine immunodeficiency-like virus.Virology 175:391-409.

13. Gourdou, I.,V. Mazarin, G. Quert, N. Sauze, and R. Vigne. 1989. The open reading frame S ofvisna virus genome is a trans-activatinggene.Virology 171:170-178.

14. Greene,W.C. 1991. The molecularbiologyof human

immuno-deficiency

virus type 1infection.N.Engl.J. Med.324:308-317. 15. Guatelli, J. C., T. R. Gingeras, and D. D. Richman. 1990. Alternativespliceacceptorutilization duringhuman

immunode-ficiency

virus type 1 infection of cultured cells. J. Virol. 64:4093-4098.

16. Guyader, M.,M.Emerman,P.Sonigo,F.Clavel, L.Montagnier, andM. Alizon. 1987. Genomeorganizationandtransactivation ofthe humanimmunodeficiencyvirus type 2. Nature (London) 326:662-669.

17. Hope,T.J.,D.McDonald,X.Huang, J. Low, and T. G. Parslow. 1990.Mutationalanalysisof the humanimmunodeficiencyvirus type 1 Rev transactivator: essential residues near the amino terminus. J. Virol. 64:5360-5366.

18. Horikami, S. M., and S. A. Moyer. 1991. Synthesis of leader RNA and editing of the P mRNA during transcription by

purified measlesvirus. J.Virol. 65:5342-5347.

19. Kalinski,H.,A.Yaniv,P.Mashiah,T.Miki, S. R.Tronick,and A. Gazit. 1991. Rev transcripts ofcaprine arthritis encephalitis virus. Virology 183:786-792.

20. Kawakami, T.,L.Sherman, J.E.Dahlberg,A.Gazit,A.Yaniv, S. R.Tronick, and S. A. Aaronson. 1987. Nucleotide sequence analysis of the integrated form of the genome ofequine infec-tiousanemiavirus. Virology 158:300-312.

21. Kjems, J., A. D. Frankel, and P. A. Sharp. 1991. Specific

regulation ofmRNAsplicingin vitroby apeptide from HIV-1 Rev. Cell 67:169-178.

22. Kozak, M. 1987. Effects of intercistronic lengthon theefficiency

of reinitiationby eucaryotic ribosomes. Mol. Cell. Biol. 7:3438-3445.

23. Kozak, M. 1989. Context effects and inefficient initiation at non-AUG codons in eucaryotic cell-free translation systems. Mol. Cell. Biol. 9:5073-5080.

24. Malim, M. H., S. Bohnlein, J. Hauber, and B.R. Cullen. 1989. Functional dissection of the HIV-1 Revtrans-activator deriva-tion ofatrans-dominant repressor of Rev function. Cell 58:205-214.

25. Malim, M. H., J. Hauber, S.-Y.Le, J. V. Maizel, and B. R. Cullen. 1989. The HIV-1 rev trans-activator acts through a structural target sequence to activate nuclear export of

un-spliced viralmRNA. Nature (London)338:254-257.

26. Malim, M. H., D. F. McCarn, L. S. Tiley, and B. R. Cullen. 1991. Mutational definition of the human immunodeficiency virus type 1 Rev activation domain. J. Virol. 65:4248-4254. 27. Malim, M. H., L. S. Tiley, D. F. McCarn, J. R. Rusche, J.

Hauber, and B. R. Cullen. 1990. HIV-1 structural gene expres-sion requires binding of the Rev trans-activator to its RNA target sequence. Cell 60:675-683.

28. Mazarin, V., I. Gourdou, G. Querat, N. Sauze, G.Audoly, C. Vitu, P. Russo, C. Rousselot, P. Filippi, and R. Vigne. 1990. Subcellular localization of rev-gene product in visna virus-infected cells. Virology 178:305-310.

29. Mazarin, V., I. Gourdou, G.

Querat,

N. Sauze, andR. Vigne. 1988. Genetic structure and function of an early transcript of visna virus. J. Virol. 62:4813-4818.

30. Mild, T.,T.Matsui, M. A. Heidaran, and S. A. Aaronson. 1989. An efficient directional cloning system to construct cDNA libraries containing full-length inserts at high frequency. Gene 83:137-146.

31. Miller, M. A., R. F.

Garry,

J. M. Jaynes, and R. C. Montelaro. 1991. A structural correlation between lentivirus transmem-brane proteinsandnatural cytolytic peptides. AIDS Res.Hum. Retroviruses 7:511-519.

32. Neumann, J. R., C. A. Morency, and K. 0. Russian. 1987. A novel rapid assay for chloramphenicol acetyltransferase gene expression. BioTechniques 5:444-447.

33. Noiman, S., A. Gazit, 0. Tori, L. Sherman, T. Miki, S. R. Tronick, and A.Yaniv. 1990. Identification of sequences encod-ing the equine infectious anemia virus tat gene. Virology 176: 280-288.

34. Noiman, S., A. Yaniv, L. Sherman, S. R. Tronick, and A. Gazit. 1990.Pattern of transcription of the genome of equine infectious anemia virus. J. Virol. 64:1839-1843.

35. Noiman, S., A. Yaniv, T. Tsach, T.Miki,S. R. Tronick, and A. Gazit. 1991. The Tat protein of equine infectious anemia virus is encoded by at least three types of transcripts. Virology 184:521-530.

36. Oberste, M. S., J. D. Greenwood, and M. A. Gonda. 1991. Analysis of the transcriptionpattern and mapping of the puta-tive rev and env splice junctions of bovine immunodeficiency-like virus. J. Virol. 65:3932-3937.

37. Olsen, H. S., P. Nelbock, A. W. Cochrane, and C. A. Rosen. 1990.Secondary structure is the major determinant for interac-tion of HIV rev protein with RNA. Science 247:845-848. 38. Perkins, A., A. W. Cochrane, S. M. Ruben, and C. A. Rosen.

1989. Structural and functional characterization of the human immunodeficiency virus Rev protein. J. Acquired Immune

on November 9, 2019 by guest

http://jvi.asm.org/

fic. Syndr.2:256-263.

39. Querat, G., G.Audoly,P. Sonigo, and R. Vigne. 1990. Nucle-otide sequence

analysis

of SA-OMVV, a visna-related ovine lentivirus: phylogenetic history of lentiviruses. Virology 175: 434 447.

40. Rice, N. R., L. E. Henderson, R. C. Sowder, T. D. Copeland, S. Oroszlan, and J. F. Edwards.1990. Synthesis andprocessing of the transmembrane envelope protein of equineinfectious ane-mia virus.J.Virol. 64:3770-3778.

41. Robert-Guroff, M., M.Popovic,S. Gartner, P.Markham, R. C. Gallo,and M.S. Reitz.1990. Structureandexpression oftat-, rev-, andnef-specific transcripts of human immunodeficiency virustype1in infectedlymphocytes andmacrophages. J. Virol. 64:3391-3398.

42. Rushlow, K., K. Olsen, G. Stiegler, S. L. Oayne, R. C. Mon-telaro,andC. J. Issel. 1986. Lentivirusgenomic organization: the complete nucleotide sequence of the env gene region of equineinfectious anemiavirus. Virology 155:309-321. 43. Saltarelli, M., G. Querat, D. A. M. Konings, R. Vigne, and J. E.

Clements. 1990.Nucleotidesequenceandtranscriptional analy-sis of molecular clones of CAEVwhich generate infectious virus.Virology179:347-364.

44. Sargan,D.R., andI.D. Bennet.1989. Atranscriptionalmapof visnavirus: definition ofthesecond intronstructuresuggestsa

rev-likegeneproduct. J. Gen.Virol. 70:1995-2006.

45. Sherman, L., A.Gazit,A.Yaniv,T.Kawakami, J.E.Dahlberg, andS.R. Tronick. 1988. Localization ofsequencesresponsible

for trans-activation of theequine infectious anemia virus long

terminalrepeat.J.Virol. 62:120-126.

46. Spence, A. M.,P. C. Sheppard, J. R.Davie, Y. Matou, N. Nishi, W. L. McKeenan, J. G. Dodd, and R. J. Matusik. 1989. Regulation of a bifunctional mRNA results in synthesis of secreted and nuclear probasin. Proc. Natl. Acad. Sci. USA 86:7843-7847.

47. Stephens,R.M., D. Derse, and N. R. Rice. 1990. Cloning and characterization ofcDNAs encoding equineinfectiousanemia virus Tat and putative Revproteins. J. Virol.64:3716-3725. 48. Tiley, L.S., P. H. Brown,S.-Y. Le, J. V. Maizel, J. E. Clements,

and B. R. Cullen. 1990. Visna virus encodesa post-transcrip-tional regulator of viral structuralgene expression.Proc.Natl. Acad.Sci.USA87:7497-7501.

49. Unger, R. E., M. W. Stout, and P. A. Luciw. 1991. Simian immunodeficiency virus(SIVmac) exhibits complex splicing for tat, revandenvmRNA.Virology 182:177-185.

50. Venkatesh, L. K., and G. Chinnadura. 1990. Mutants in a conserved region near the carboxy-terminus of HIV-1 Rev identify functionally important residues and exhibitadominant negative phenotype. Virology 178:327-330.

51. Venkatesh, L. K., S. Mohammad, and G. Chinnadura. 1990. Functional domains oftheHIV-1 rev gene required for trans-regulation and subcellular localization. Virology 176:39-47. 52. Viglianti,G.A., P. L.Sharma,and J. I. Mullins.1990.Simian

immunodeficiency virus displays complex patterns of RNA splicing.J. Virol. 64:4207-4216.

53. Zapp, M. L., and M. R. Green.1989. SequencespecificRNA binding by the HIV-1 Revprotein. Nature(London) 342:714-716.

on November 9, 2019 by guest

http://jvi.asm.org/