6
Vol. 62, No. 12 Genetic Structure and Function of an Early Transcript of Visna Virus VERONIQUE MAZARIN, ISABELLE GOURDOU, GILLES QUERAT, NICOLE SAUZE, AND ROBERT VIGNE* Laboratoire de Virologie, Faculte de Medecine Nord, 13326 Marseille Cedex 15, France Received 9 May 1988/Accepted 15 August 1988 During the early step of the lytic cycle, visna provirus is first transcribed into two small multispliced mRNAs of 1.6 and 1.2 kilobases which may encode factors regulating the replication of visna virus (R. Vigne, V. Barban, G. Querat, V. Mazarin, I. Gourdou, and N. Sauze, Virology 161:218-227, 1987). By cDNA cloning and nucleotide sequencing, we determined that the 1.2-kilobase mRNA is 1,174 nucleotides long without the 3'-polyadenylated tail and is composed of four exons, two of which originated from the 5' and 3' ends, respectively, of the env gene region. Two overlapping open reading frames are present in each of these two exons. They were translated in vitro and gave rise to three proteins, two of 19 and 17 kilodaltons, termed VEP1, and one of 16.5 kilodaltons, termed STM. Only the VEP1 proteins were recognized by a hyperimmune anti-visna virus serum of infected sheep. Transient-expression assays performed in eucaryotic cells demon- strated that the cDNA clone described here has a trans-acting effect on transcription of the visna virus genes. Visna virus, one of the earliest isolates of animal lentivi- ruses, was discovered in sheep affected by degenerative progressive encephalitis and pneumonia (18). Although its pathogenesis has been well studied in the last decades (8), the biological importance of visna virus has recently been emphasized by demonstration of the close relatedness of visna virus to human immunodeficiency virus, the causative agent of acquired immunodeficiency syndrome (7). Like the other retrovirus genomes, the visna virus genome contains three open reading frames (ORFs), gag, pol, and env, which encode capsid proteins, reverse transcriptase, and envelope protein(s), respectively. The visna virus genome contains at least two other ORFs, previously designated orfQ and orfS, which are probably implicated in the regulation of viral expression (20). Interestingly, we have recently obtained evidence of regulation of visna virus gene transcription in infected ovine cells, with the observation of preferential early expression of two small multispliced mRNAs of 1.6 and 1.2 kilobases (kb) at 24 h postinfection (p.i.) and subsequent expression of small and large mRNAs at 72 h p.i. (21). The large transcripts are the genomic RNA of 9.4 kb, which is translated into the structural gag and pol products, and three singly spliced mRNAs of 4.8, 4.3, and 3.7 kb, which are probably translated into the env products and the putative orfQ product (5, 21). The small multispliced 1.6- and 1.2-kb mRNAs may be translated into products that regulate the accumulation of the late viral products, as recently observed for the tat and art-trs proteins of human immuno- deficiency virus (4). To determine the exact genetic content of the early transcripts and their functions in the virus cell cycle, we cloned, sequenced, and expressed in vitro the cDNA corresponding to the 1.2-kb mRNA. The structure and role of the first lentivirus early gene, VEGI, are pro- posed. American ovine fetal trachea cells infected for 24 h with visna virus strain K1514 were the source of mRNA (21). Cytoplasmic polyadenylated RNA was isolated by oligo(dT)- cellulose column chromatography, and small mRNAs of 1 to * Corresponding author. 2 kb were purified by centrifugation in a 5 to 20% sucrose gradient. Double-stranded cDNA was synthesized with the Amersham kit according to the instructions of the manufac- turer, ligated to EcoRI linkers, and cloned into the EcoRI sites of bacteriophage lambda gtlO DNA (12). After DNA packaging, the recombinant plaques were screened by in situ hybridization with U5-leader and U3 probes. The selected cDNAs were recloned into an M13-like phage (Phagescript; Stratagene), sequentially deleted by exonuclease III and mung bean nuclease, and then sequenced by the termination chain method of Sanger et al. (17). The selected cDNA (termed 16.1 cDNA) was 783 nucleo- tides long (Fig. 1A). In comparison with the genetic structure expected for small transcripts (21), the 16.1 cDNA clone lacked 27 nucleotides at its 5' end and 364 nucleotides at its 3' end. Computer-generated nucleotide sequence analysis of the R region revealed many secondary structures that might interfere during cDNA synthesis by blocking the reverse transcriptase reaction. Truncation of the 364 nucleotides suggested that thq reverse transcriptase reaction was primed at the A-rich stretch upstream of PPT2 instead of at the 3' poly(A) tail. This hypothesis was strengthened by the pres- ence of a G-to-A base substitution, which took place at position 807 of the 16.1. cDNA clone sequence. Therefore, the complete cDNA should be 1,174 nucleotides long and should correspond to a polyadenylated mRNA of 1,374 nucleotides. To investigate the nature of the 16.1 cDNA, we recon- structed its defective 3' end with a HindIII-SstI fragment of the visna provirus lambda vis 109 clone (20). The resulting sequence, named 1.4 cDNA, spans the theoretical sequence described in Fig. 1A from position 28 to position 1100 (SstI site). The 1.4 cDNA was transcribed into the Phagescript system under the direction of the T7 promoter and gave rise to a 1,165-nucleotide RNA. This novel RNA was analyzed in parallel with polyadenylated RNA from 24-h visna virus- infected cells by Northern (RNA) blotting with a U5-leader probe. The 1,165-nucleotide RNA migrated slightly faster than did the smallest mRNA of visna virus, which we had previously termed 1.2-kb mRNA (Fig. 1B). More important, 4813 JOURNAL OF VIROLOGY, Dec. 1988, p. 4813-4818 0022-538X/88/124813-06$02.00/0 Copyright © 1988, American Society for Microbiology on April 12, 2019 by guest http://jvi.asm.org/ Downloaded from

Genetic Structure and Function Early Transcript Visna Virus

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Genetic Structure and Function Early Transcript Visna Virus

Vol. 62, No. 12

Genetic Structure and Function of an Early Transcriptof Visna Virus

VERONIQUE MAZARIN, ISABELLE GOURDOU, GILLES QUERAT, NICOLE SAUZE, AND ROBERT VIGNE*

Laboratoire de Virologie, Faculte de Medecine Nord, 13326 Marseille Cedex 15, France

Received 9 May 1988/Accepted 15 August 1988

During the early step of the lytic cycle, visna provirus is first transcribed into two small multispliced mRNAsof 1.6 and 1.2 kilobases which may encode factors regulating the replication of visna virus (R. Vigne, V.Barban, G. Querat, V. Mazarin, I. Gourdou, and N. Sauze, Virology 161:218-227, 1987). By cDNA cloningand nucleotide sequencing, we determined that the 1.2-kilobase mRNA is 1,174 nucleotides long without the3'-polyadenylated tail and is composed of four exons, two of which originated from the 5' and 3' ends,respectively, of the env gene region. Two overlapping open reading frames are present in each of these twoexons. They were translated in vitro and gave rise to three proteins, two of 19 and 17 kilodaltons, termed VEP1,and one of 16.5 kilodaltons, termed STM. Only the VEP1 proteins were recognized by a hyperimmuneanti-visna virus serum of infected sheep. Transient-expression assays performed in eucaryotic cells demon-strated that the cDNA clone described here has a trans-acting effect on transcription of the visna virus genes.

Visna virus, one of the earliest isolates of animal lentivi-ruses, was discovered in sheep affected by degenerativeprogressive encephalitis and pneumonia (18). Although itspathogenesis has been well studied in the last decades (8),the biological importance of visna virus has recently beenemphasized by demonstration of the close relatedness ofvisna virus to human immunodeficiency virus, the causativeagent of acquired immunodeficiency syndrome (7). Like theother retrovirus genomes, the visna virus genome containsthree open reading frames (ORFs), gag, pol, and env, whichencode capsid proteins, reverse transcriptase, and envelopeprotein(s), respectively. The visna virus genome contains atleast two other ORFs, previously designated orfQ and orfS,which are probably implicated in the regulation of viralexpression (20). Interestingly, we have recently obtainedevidence of regulation of visna virus gene transcription ininfected ovine cells, with the observation of preferentialearly expression of two small multispliced mRNAs of 1.6and 1.2 kilobases (kb) at 24 h postinfection (p.i.) andsubsequent expression of small and large mRNAs at 72 h p.i.(21). The large transcripts are the genomic RNA of 9.4 kb,which is translated into the structural gag and pol products,and three singly spliced mRNAs of 4.8, 4.3, and 3.7 kb,which are probably translated into the env products and theputative orfQ product (5, 21). The small multispliced 1.6- and1.2-kb mRNAs may be translated into products that regulatethe accumulation of the late viral products, as recentlyobserved for the tat and art-trs proteins of human immuno-deficiency virus (4). To determine the exact genetic contentof the early transcripts and their functions in the virus cellcycle, we cloned, sequenced, and expressed in vitro thecDNA corresponding to the 1.2-kb mRNA. The structureand role of the first lentivirus early gene, VEGI, are pro-posed.American ovine fetal trachea cells infected for 24 h with

visna virus strain K1514 were the source of mRNA (21).Cytoplasmic polyadenylated RNA was isolated by oligo(dT)-cellulose column chromatography, and small mRNAs of 1 to

* Corresponding author.

2 kb were purified by centrifugation in a 5 to 20% sucrose

gradient. Double-stranded cDNA was synthesized with theAmersham kit according to the instructions of the manufac-turer, ligated to EcoRI linkers, and cloned into the EcoRIsites of bacteriophage lambda gtlO DNA (12). After DNApackaging, the recombinant plaques were screened by in situhybridization with U5-leader and U3 probes. The selectedcDNAs were recloned into an M13-like phage (Phagescript;Stratagene), sequentially deleted by exonuclease III andmung bean nuclease, and then sequenced by the terminationchain method of Sanger et al. (17).The selected cDNA (termed 16.1 cDNA) was 783 nucleo-

tides long (Fig. 1A). In comparison with the genetic structureexpected for small transcripts (21), the 16.1 cDNA clonelacked 27 nucleotides at its 5' end and 364 nucleotides at its3' end. Computer-generated nucleotide sequence analysis ofthe R region revealed many secondary structures that mightinterfere during cDNA synthesis by blocking the reversetranscriptase reaction. Truncation of the 364 nucleotidessuggested that thq reverse transcriptase reaction was primedat the A-rich stretch upstream of PPT2 instead of at the 3'poly(A) tail. This hypothesis was strengthened by the pres-ence of a G-to-A base substitution, which took place atposition 807 of the 16.1. cDNA clone sequence. Therefore,the complete cDNA should be 1,174 nucleotides long andshould correspond to a polyadenylated mRNA of 1,374nucleotides.To investigate the nature of the 16.1 cDNA, we recon-

structed its defective 3' end with a HindIII-SstI fragment ofthe visna provirus lambda vis 109 clone (20). The resultingsequence, named 1.4 cDNA, spans the theoretical sequencedescribed in Fig. 1A from position 28 to position 1100 (SstIsite). The 1.4 cDNA was transcribed into the Phagescriptsystem under the direction of the T7 promoter and gave riseto a 1,165-nucleotide RNA. This novel RNA was analyzed inparallel with polyadenylated RNA from 24-h visna virus-infected cells by Northern (RNA) blotting with a U5-leaderprobe. The 1,165-nucleotide RNA migrated slightly fasterthan did the smallest mRNA of visna virus, which we hadpreviously termed 1.2-kb mRNA (Fig. 1B). More important,

4813

JOURNAL OF VIROLOGY, Dec. 1988, p. 4813-48180022-538X/88/124813-06$02.00/0Copyright © 1988, American Society for Microbiology

on April 12, 2019 by guest

http://jvi.asm.org/

Dow

nloaded from

Page 2: Genetic Structure and Function Early Transcript Visna Virus

4814 NOTES

666AAAA6CA6A6T6CTGTG6AGA6CTC6AA6GAAAA6ATCTCCG66CCTCTCCTGCCTGCCTGAAAAGC 70L

TCAATAAAGGA6TTGCT6ATATCT6AGCTTGCTGGTTATTATC66GATTCGTTACTAATTCC6T6CAAC 140

ACC66AGC66ATCTCGCA6CTGGC6CCCAACGTGGG6CTCGACAAA6AATCAGAAGAAAAAT6AGA6TTA 210

TGGGACCAC66ACGCTGCTCCTGTGAGGAC6GCjA66AGAGTAACGGACACGGACAAAAAGTGAAAGAAA 280

GCTTC6666ACGCCT6AAGTAAGAACAAGAAAAAGAGGGCATCCA66AGAGTG6CAAGGACCAACACAGC 350303'-4843 4888 5946

CTGTCTAGGATGGCCAGCAAGAAAGTAAGCCAAGCAGAACAACGTGGAGAGACATGGAGCCCCCACTCC 420

G6AAMACAT6GAATCAA6TATTACA66AGCTAGTAAAAAGACAACAACAGGATA6AG6A6AMCAGCAGGG 490

ACTGGTATCAG6CTTACAAGCAAGTAAAGCAGATCAGATATACACAGGTAACAGTGGTGATAGAAGCACC 5606097A8530

6G6TGAATT6GA66AAAAACAAAAAAGAAAC6GGGATGGTACAAAT6GCTGCGCAAGCTTA6AGCAC6AG 630

AGAAGAACATCCCATCGCAGTTTTATCCAGATATGGAGAGCAACATGGTGGGCATGGAAAACCTCACCTT 700

66AGACACAACT6GAGGACAATGCCCTATATAACCCTGCTACCCATATTGGTGATATGGCAATGGATGGA 770

A6AGAATGGATGGAATGGAGAGMATCAGCACAAAAAAMAAAAGAAAGGGTGGACTGTCAGGACAGAGAA 840

CAAAT6CCTACCCTGGAAAATGACTATGTAGAGTTATAGGAAGGTCAT6TCACTGTTACCA6AAMTCATA 910

GTCAG6AT6ACACAGCAAATGTAACCAGTTACCAGAAATCATAGTCAGGAT$ACACAGCAAATGTAACCG 980

CAAGTTCTGCrTTTTTGC6CTAAGTCAT6TAGCAGCTGArGCTTGAGTCATAACCGCAGATGTAAACAAG 1050

TT6CCTATATAA6CCGCTTGCTAGCTGGGGAAAAGCAGAGTGCTTTGGAGAGCTCGAMGGAAAGAGTCTC 1120

CG6GCCTCTCCT6CCT6CCTGAAAA6CTCAATAAGAGTT6GCTGATATCTGA 1174PoLYASIGNAL

C.

1 2

- 9.4 kb

4.8 kb-4.3kb

3.7 kb

1.6kb

_ 1.2kb40

RU5L U3 R

_ GCAE,,,___ P^Ot-L_ -Qj ENV ,,&,,,,,4kb

4 exon Il 2.:: 3 exon4

5s56 w 068'RNA

1 AVEP1

13-A ---- STM

As

Be

J. VIROL.

135 AA...

on April 12, 2019 by guest

http://jvi.asm.org/

Dow

nloaded from

Page 3: Genetic Structure and Function Early Transcript Visna Virus

NOTES 4815

5 10 15 20 25 36H A S K E S K P S R T T U R 0 h E P P L R E T U N 0 VL 0 E

35 46 45 1s 55 66L V K R Q 0 0 E E E Q 0 6 L V S 6 L 0 A S K AD 0 1 Y T 6

65 76 75 o6 65 96

N S 6 0 R S T 6 6 1 6 6 K T K K K R 6 U Y K W L R K L RA R

95 160 165 lie I15 126E K N I P S 0 F Y P 0 h E S N M V 6 E N L T L E T Q L E 0

125 130 135 146 145 156

N A L Y N P A T H I 6 0 A 0 6 R E W E W R E S A Q K K

155 160 165K R K 6 6 L S 6 Q R T N A Y P 6 K

B.S Is I5 26 25 30

M K K N SR 0 Y 0 AY K 0 VK 0 I R Y T Q V T V V I E A P

35 40 45 56 55 66V E L E E K 0 K R N 6 0 6 T N 6 C A S L E H E R R T S H RS

65 76 7S 85 96

F I 0 1 U R A T U WiA K T S P U R H N W R t P Y I T L L

95 166 165 116 115 120P I L V I W 0 W E E N 6 U N 6 E N Q H K K K K E R V 0 CQ

125 136 1350 R E 0 P T L E N 0 Y V E L

exon 3 exon 4

AAL.

FIG. 2. Deduced amino acid sequences and corresponding hydropathic profiles of VEP1 (A) and STM (B). The boundaries between thetwo encoding exons are indicated by a vertical line; amino acids described in the text are underlined (K, R, and W).

it was not detected with the probes that were specific to theother viral mRNAs (data not shown). These results suggestthat the cloned 16.1 cDNA corresponds to the smallest earlymRNA, herein termed 1.4-kb (its exact length) mRNA.The genetic organization of this transcript is presented in

Fig. 1C. The transcript is composed of four exons that resultfrom splicing of the 9.4-kb genomic mRNA at positions 303,4888, and 6097 for donor sites and at positions 4843, 5946,and 8530 for acceptor sites.The 1.4-kb mRNA exon 1 (positions 1 to 303 on the 9.4-kb

mRNA sequence) consisted of the R, U5, and leader regions,which are probably essential for the stability or translationalefficiency of mRNAs.Exon 2 (positions 4843 to 4888) corresponds to a small

region of the pol 3'-end gene. According to Northern blotanalysis of small transcripts with probes representative ofthe 3' end of pol, these sequences do not seem to be presentin all of the small transcripts of visna virus-infected cells(data not shown).Exons 3 (positions 5946 to 6097) and 4 (positions 8530 to

9202) of the 1.4-kb mRNA consist of the 5' and 3' regions,respectively, of the env gene (Fig. 1C). Computer-generatedsequence analysis revealed that exons 3 and 4 produce twoalternate ORFs. The longest ORF, initiated in exon 3 at theATG1 codon (position 5956), could encode a 167-amino-acidprotein. Since ATG1 is also the putative env protein initia-tor, this product will contain 48 amino acids also found in theenv protein precursor, the remaining 119 amino acids beingtranslated from exon 4 into an alternative reading frame.Another ATG codon (ATG2), located at position 6001 in the

ORF defined by ATG1, can initiate a 152-amino-acid proteinin which the 15 N-terminal residues from the first ATG are

missing. The 1.4-kb mRNA product, initiated at the ATG1(or ATG2) codon, is a new protein that we have named visnaearly protein 1 (VEP1).The deduced amino acid sequence and corresponding

hydropathic profile of VEP1 are shown in Fig. 2A. VEP1does not exhibit distinctive hydrophobic regions, but inter-estingly, the domain encoded by exon 3 contains a stretch ofbasic amino acids (positions 73 to 92) reminiscent of a clusterof lysine-arginine residues observed in the art-trs product ofhuman immunodeficiency virus (19) and in the hinge domainof nucleic acid-binding proteins such as glucocorticoid andestrogen receptors (1). The 1.4-kb mRNA exon 3 showsanother ATG codon, termed ATG3, which is located atposition 6068. This codon presents a typical codon start site(13) and defines another ORF, which can encode a 135-amino-acid protein (Fig. 1C). This putative protein, whichwe termed short transmembrane protein (STM), is encodedmainly by exon 4, which in fact corresponded to the last 125amino acids of the C-terminal region of the envelope trans-membrane protein.To characterize the proteins encoded by the 1.4-kb

mRNA, the 1.4 cDNA subcloned into the Phagescript vector(see above) was transcribed by T7 polymerase, and theresulting transcript was translated in a rabbit reticulocytelysate (Promega Biotec kit) in the presence of [35S]methi-onine. The labeled polypeptides were electrophoresed insodium dodecyl sulfate-15% polyacrylamide gels and fluo-rographed. Representative results are shown in Fig. 3A, lane

FIG. 1. Nucleotide sequence (A), Northern blot analysis (B), and genetic structure (C) of the 1.4-kb mRNA. (A) The 16.1 cDNA, spanningpositions 28 to 810, is enclosed by brackets. Positions of the splice sites of the genomic RNA were indicated on the cDNA sequence (20). Thesequence determined was identical to that of the corresponding portions of the provirus genome reported previously (20) except for thefollowing base replacements: T to G at position 473 and A to G at position 807. (B) The 1.4 cDNA was obtained by reconstructing the 16.1cDNA defective 3' end with a HindIII-SstI fragment of visna provirus lambda vis 109. The resulting sequence was transcribed into thePhagescript system and gave rise to the 1,165-nucleotide RNA shown in lane 1. Poly(A)-containing mRNAs were isolated from Americanovine fetal cells infected with visna virus for 24 h (lane 2). RNAs were fractionated by formaldehyde-agarose gel electrophoresis, and viralRNAs were detected with a 32P-labeled visna virus U5-leader probe. (C) The location of the four exons in the 1.4-kb mRNA is shown at the

positions of the splice sites on the 9.4-kb genomic RNA. The two resulting ORFs, encoding the two VEP1 proteins and STM, are shown atthe positions on the genomic RNA of the three ATGs initiating the three proteins.

A.

exon3 e Kon 4

-Ifrqr' 'rqlu.Mmw.-$ -rM Li

VOL. 62, 1988

j hk a I

on April 12, 2019 by guest

http://jvi.asm.org/

Dow

nloaded from

Page 4: Genetic Structure and Function Early Transcript Visna Virus

4816 NOTES

A 1 2 3 4

U_~_SM

B 2 3 4 A. p KVEGi

- 19 kDa17 kDa

- 16.5 kDa

FIG. 3. In vitro translation of the 1.4-kb mRNA. (A) Translationproducts. The 1,165-nucleotide RNA was translated in a rabbitreticulocyte lysate in the presence of [35S]methionine (lane 3) or

[35S]cysteine (lane 4). Lanes 1 and 2 are cysteine and methioninetranslation controls, respectively, without RNA. The labeled prod-ucts were electrophoresed in sodium dodecyl sulfate-15% poly-acrylamide gels and fluorographed. (B) Immunoprecipitation of thein vitro-translated products. [35S]methionine-labeled polypeptides(lane 1) were incubated with anti-visna virus sheep serum (lane 2) orwith nonimmune sheep serum (lane 3). The immune precipitateswere concentrated on S. aureus Formalin-fixed cells. Lane 4, Invitro translation products nonimmunologically bound to the bacte-ria. The immune precipitates were analyzed on sodium dodecylsulfate-15% polyacrylamide gels and fluorographed.

3. Although multiple bands are visible, our attention wasfocused on the three largest translated polypeptides, whichhad apparent molecular sizes of 19, 17, and 16.5 kilodaltons(kDa). The first two molecular sizes corresponded to themolecular sizes calculated for two proteins related to VEP1initiated at ATG1 and ATG2, respectively; the third corre-sponded to a protein related to STM that was initiated atATG3. We distinguished between VEPl- and STM-relatedpolypeptides by the fact that STM could be distinctly labeledwith [35S]cysteine. Cysteine was specifically incorporatedinto the 16.5-kDa polypeptide (Fig. 3A, lane 4), whichsuggested its relatedness to STM. In contrast, the two largerproteins (19 and 17 kDa), which were not labeled by[35S]cysteine, were probably related to VEP1. To demon-strate that these proteins were produced in vivo, the[35S]methionine-labeled products were tested for immunereactivity to a high-titer antiserum that was raised in sheepexperimentally infected with visna virus. The VEPl-related19- and 17-kDa proteins were specifically recognized by theanti-visna virus sheep serum but not by a nonimmune sheepserum (Fig. 3B, lanes 2 and 3). It is noteworthy that theimmune and nonimmune sera reacted similarly with the16.5-kDa STM product. Moreover, this polypeptide reactednonspecifically with Staphylococcus aureus Formalin-fixedbacteria (Fig. 3B, lane 4). We therefore conclude that onlyVEP1 polypeptides were immunologically recognized byinfected animals.To demonstrate the functional properties of the 1.4 cDNA

clone, we performed transient-expression assays in which a

1.4 cDNA-expressing vector (pKVEG1) was cotransfectedwith a plasmid (pVLTR) carrying the visna virus long-terminal-repeat (LTR) sequences fused to the beta-globingene. pKVEG1 was obtained by inserting the 1.4 cDNA intothe unique EcoRI site ofpKCR3 between the simian virus 40early gene promoter and the splicing and polyadenylationsites of the beta-globin gene (Fig. 4A) (14). pVEG1 couldthus transiently express the 1.4 cDNA when introduced inCOS-CV1 cells. Construction of pVLTR will be describedelsewhere (I. Gourdou et al., manuscript in preparation).

--- 1.4 cDNA.

* GLOBIN GENE.

&q SV40 ori, EARLY PROMOTERAND POLYADENYLATIONSITES.

LTR SEQUENCES.* B-GLOBIN GENE.

B.

1 2 3 4 5

FIG. 4. trans-Activation effect of the 1.4 cDNA products. (A)Recombinant plasmids used for cotransfection assays. pKVEG1was constructed by ligating the 1.4 cDNA into the EcoRl site ofpKCR3; it contains the ori sequences, the early gene promoter andpolyadenylation sites of simian virus 40 (SV40), and part of therabbit beta-globin gene (part of exons, 2 and 3). pVLTR was

prepared as follows: after restoration of the blunt end, a HindIIlfragment (559 nucleotides) containing the complete visna virus LTRwas subcloned into the unique SmaI site of pMT4, upstream of thecomplete rabbit beta-globin gene (exons 1, 2, and 3). The geneticstructures of the spliced and expected polyadenylated transcriptsare shown. Abbreviations: E, EcoRI; S, SmaI; B, BamHI; A,poly(A) site. (B) Beta-globin mRNA accumulation. COS-CV1 cells(107) were exposed to 2 pmol of pVLTR and to 0.5 or 2 pmol ofthe plasmid to be analyzed. Transfections were performed by thecalcium phosphate precipitation procedure, and total cytoplasmicRNA was extracted 48 h posttransfection. mRNA was isolated on

oligo(dT)-cellulose as follows: 2 ,ug was loaded in each well of a

formaldehyde-agarose gel and analyzed by the Northern blot tech-nique, using a 32P-labeled probe specific for the pVLTR beta-globingene (exon 1). Lanes: 1, 2 pmol of pVLTR; 2 and 3, 2 pmol ofpVLTR plus 0.5 and 2 pmol, respectively, of the control plasmid,pKCR3; 4 and 5, 2 pmol of pVLTR plus 0.5 and 2 pmol, respec-tively, of pKVEG1.

Briefly, it was obtained by subcloning the complete LTRsequences into the unique SmaI site of pMT4 (2), which islocated upstream of a complete rabbit beta-globin gene.Thus, the LTR sequences promote the expression of thebeta-globin gene when pVLTR is introduced in cells express-ing the appropriate transcription factors. Cotransfectionassays were performed in COS-CV1 cells (6) by the calciumphosphate precipitation procedure (22). The cells were ex-

posed to constant amounts of pVLTR and to variousamounts of plasmid DNA to be analyzed and processed formRNA analysis 48 h after the transfection experiment. TheLTR-directed transcripts were visualized after Northern blotanalysis by hybridization with a probe specific for thepVLTR beta-globin gene (Fig. 4B). Comparisons of theresults obtained when pVLTR was transfected alone (lane 1)with results obtained when pVLTR was cotransfected with0.5 or 2 pmol of the control plasmid pKCR3 (lanes 2 or 3,respectively) indicated that the steady-state level of beta-globin mRNA remained constant. In contrast, the steady-

pV LTR

J. VIROL.

on April 12, 2019 by guest

http://jvi.asm.org/

Dow

nloaded from

Page 5: Genetic Structure and Function Early Transcript Visna Virus

NOTES 4817

state level of the beta-globin mRNA increased in the pres-ence of increased amounts of pKVEG1, indicating that thetranscription reaction directed by the LTR sequences wasenhanced by the pKVEGI products. This effect was mea-sured by scanning the X-ray film. The results obtained forthe experiment described above showed that the LTR-directed transcription was enhanced 8.7-fold when the assaywas performed with 2 pmol of pKVEG1. Other experimentsperformed under the same conditions resulted in higherenhancement values (up to 30-fold); this variability probablyreflected differences in cell culture conditions. Nevertheless,all of the experiments we performed led to the same impor-tant conclusion: the 1.4 cDNA clone was biologically activeand encoded products which, directly or indirectly, trans-activated the visna virus LTR sequences.We have described here the structure and function of an

early-expressed visna virus gene, which we have namedVEGI. The corresponding 1.4-kb transcript, previouslytermed 1.2-kb mRNA (21), is the smallest transcript and,with a second small mRNA of 1.6 kb, represents the twoearly transcripts present in cells at 24 h p.i. The largesttranscripts (9.4, 4.8, 4.3, and 3.7 kb) were detected as majortranscripts later in the lytic cycle, between 48 and 72 h p.i.By cDNA cloning and nucleotide sequencing, we deter-mined that the 1.4-kb mRNA is 1,174 nucleotides longwithout the 3'-polyadenylated tail and is composed of fourexons. Exon 1 (R-U5-leader) and exon 2 (3' end of the polgene) are noncoding, but the last two exons (exons 3 and 4)encode with the three ATGs present in exon 3.

It was of particular interest to observe that, as is the casefor human immunodeficiency virus transcripts (15), the shortnoncoding exon of the 3' end ofpol (exon 2) was acquired bythe visna virus 1.4-kb mRNA. Moreover, in both virusinfection systems, this exon may be absent (I. Gourdou,personal communication) and may only slightly affect thebiological activity of the short transcripts (16).The 1.4 cDNA could be translated in rabbit reticulocyte

lysates in three proteins: VEP1 (19 and 17 kDa) and STM(16.5 kDa). VEP1 consists of two parts. The first part,encoded from the central region of the genome, correspondsto an amino acid sequence that precedes the signal peptide ofthe env protein precursor (20). The second part, encodedfrom the 3' region of the genome, corresponds to a new shortORF, distinct from the env ORF.These VEP1 products were recognized by an anti-visna

virus sheep serum, suggesting that the ORF encoding theseproducts is expressed in vivo. The nature of the 16.5-kDaSTM product, which corresponds to the putative intracellu-lar domain of the transmembrane glycoprotein, is less clear-cut. First, although this protein is a structural part of theenvelope protein, we observed no immune reactivity withthe anti-visna virus sheep serum. In contrast, we observednonspecific binding of this protein to protein A-bearingbacteria as well as to Fab and Fc regions of normal immu-noglobulins (data not shown). The nonimmunological stickyproperties of STM may be explained by its high content ofhydrophobic tryptophan residues in the C-terminal region.Second, whereas the ORF coding for VEP1 has been ob-served in all of the proviral clones of visna virus presentlysequenced (3, 20), the ATG that initiated STM (ATG3) wasnot found in these previous clones. Nevertheless, it isinteresting that this STM product corresponded structurallyto the putative X transmembrane product of simian immu-nodeficiency virus clones, which might be encoded by a

single multispliced mRNA that also encodes tat and art-trsproducts (11).

We showed in previous studies that the 1.4-kb mRNA wasan early transcript in the visna virus lytic cycle (22). Here,we have demonstrated that the corresponding 1.4 cDNAencodes products (VEP1) that are capable of activating thetranscription directed by the viral promoter contained in theLTR sequences and therefore may contribute to the trans-activation previously described in visna virus-infected cells(9, 10). Moreover, immunoprecipitation assays performedwith a polyclonal rabbit antiserum specific to VEP1 showedthat a protein related to VEP1 was present as soon as 24 hp.i. in visna virus-infected cells (V. Mazarin et al., inpreparation). We can therefore conclude that the 1.4-kbmRNA corresponds to the transcription of an early genewhich may play a significant role in the regulation of thevirus cell cycle.We thank P. Filippi and G. Burkhart for helpful discussions and J.

DeMartini for American ovine fetal trachea cells.This work was supported by grants from the Association de la

Recherche sur le Cancer (Villejuif), the Fondation de la RechercheMedicale (Paris; AIDS research funds), the Minist6re de l'EducationNationale (Mission de la Recherche), and the Institut de la Sante etde la Recherche Medicale (grant CRE 871017). V.M. and I.G. were,respectively, Fellows of the Ministere de la Recherche et del'Enseignement Superieur and of the Ligue Nationale de Luttecontre le Cancer.

LITERATURE CITED1. Adler, S., M. L. Waterman, X. He, and M. G. Rosenfeld. 1988.

Steroid receptor-mediated inhibition of rat prolactin geneexpression does not require the DNA-binding domain. Cell 52:685-695.

2. Boeuf, H., D. Zajchowski, P. Jalinot, C. Goding, B. Devaux, C.Hauss, and C. Kedinger. 1986. Intricate sequence elementsinvolved in constitutive and trans-activated expression from theadenovirus early ElA promoter region. Cancer Cells (ColdSpring Harbor) 4:203-215.

3. Braun, M. J., J. E. Clements, and M. A. Gonda. 1987. The visnavirus genome: evidence for a hypervariable site in the env geneand sequence homology among lentivirus envelope proteins. J.Virol. 61:4046-4054.

4. Chen, I. S. Y. 1986. Regulation of AIDS virus expression. Cell47:1-2.

5. Davis, J. L., S. Molineaux, and J. E. Clements. 1987. Visna virusexhibits a complex transcriptional pattern: one aspect of geneexpression shared with the acquired immunodeficiency syn-drome retrovirus. J. Virol. 61:1325-1331.

6. Gluzman, Y. 1981. SV40-transformed simian cells support thereplication of early SV40 mutants. Cell 23:175-182.

7. Gonda, M. A., F. Wong-Staal, R. C. Gallo, J. E. Clements, 0.Narayan, and R. V. Gilden. 1985. Sequence homology andmorphologic similarity of HTLV-III and visna virus, a patho-genic lentivirus. Science 227:173-177.

8. Haase, A. T. 1986. Pathogenesis of lentivirus infections. Nature(London) 322:130-136.

9. Hess, J. L., J. E. Clements, and 0. Narayan. 1985. Cis andtrans-acting transcriptional regulation of visna virus. Science229:482-485.

10. Hess, J. L., J. M. Pyper, and J. E. Clements. 1986. Nucleotidesequence and transcriptional activity of the caprine arthritis-encephalitis virus long terminal repeat. J. Virol. 60:385-393.

11. Hirsch, V., N. Riedel, and J. I. Mullins. 1987. The genomeorganization of STLV-III is similar to that of the AIDS virusexcept for a truncated transmembrane protein. Cell 49:307-319.

12. Huynh, T. V., R. A. Young, and R. W. Davis. 1985. Constructingand screening cDNA libraries in XgtlO and Xgtll. In D. Glover(ed.), DNA cloning techniques: a practical approach. IRLPress, Oxford.

13. Kozak, M. 1984. Compilation and analysis of sequences up-stream from the translational start site in eukaryotic mRNAs.Nucleic Acids Res. 12:857-872.

14. Matrisian, L. M., G. T. Bowden, P. Krieg, G. Furstenberger,

VOL. 62, 1988

on April 12, 2019 by guest

http://jvi.asm.org/

Dow

nloaded from

Page 6: Genetic Structure and Function Early Transcript Visna Virus

4818 NOTES

J. P. Briand, P. Leroy, and R. Breathnach. 1986. The mRNAcoding for the secreted protease transin is expressed more

abundantly in malignant than in benign tumors. Proc. Natl.Acad. Sci. USA 83:9413-9417.

15. Muesing, M. A., D. H. Smith, C. D. Cabradilla, C. V. Benton, L.A. Lasky, and D. J. Capon. 1985. Nucleic acid structure andexpression of the human AIDS/lymphadenopathy retrovirus.Nature (London) 313:450-458.

16. Muesing, M. A., D. H. Smith, and D. J. Capon. 1987. Regulationof mRNA accumulation by a human immunodeficiency virustrans-activator protein. Cell 48:691-701.

17. Sanger, I., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-

ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5468.

18. Sigurdsson, B., P. A. Palsson, and H. Grimsson. 1957. Visna, a

demyelinating transmissible disease of sheep. J. Neuropathol.Exp. Neurol. 16:389-403.

19. Sodroski, J., W. C. Goh, C. Rosen, A. Dayton, E. Terwilliger,and W. Haseltine. 1986. A second post-transcriptional trans-activator gene required for HTLV-III replication. Nature(London) 321:412-417.

20. Sonigo, P., M. Alizon, K. Staskus, D. Klatzmann, S. Cole, 0.

Danos, E. Retzel, P. Tiollais, A. Haase, and S. Wain-Hobson.1985. Nucleotide sequence of the visna lentivirus: relationshipto the AIDS virus. Cell 42:369-382.

21. Vigne, R., V. Barban, G. Qukrat, V. Mazarin, I. Gourdou, andN. Sauze. 1987. Transcription of visna virus during its lyticcycle: evidence for a sequential early and late gene expression.Virology 161:218-227.

22. Wigler, M., A. Pellicer, S. Silverstein, R. Axel, G. Urlaub, and L.Chasin. 1979. DNA-mediated transfer of the adenine phospho-ribosyltransferase locus into mammalian cells. Proc. Natl.Acad. Sci. USA 76:1373-1376.

J. VIROL.

on April 12, 2019 by guest

http://jvi.asm.org/

Dow

nloaded from