JOURNAL OF VIROLOGY, May 1992, p. 3026-30310022-538X/92/053026-06$02.00/0Copyright C 1992, American Society for Microbiology

Spleen Necrosis Virus, an Avian Immunosuppressive Retrovirus,Shares a Receptor with the Type D Simian RetrovirusesVINEET N. KEWALRAMANI,"12 ANTONITO T. PANGANIBAN,3 AND MICHAEL EMERMANl*

Program in Molecular Medicine and Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle,Washington 981041; Department of Microbiology, University of Washington, Seattle, Washington 981952; and McArdle

Laboratory for Cancer Research, University of Wisconsin, Madison, Wisconsin 537063

Received 23 December 1991/Accepted 10 February 1992

The reticuloendotheliosis viruses (REV) are a family of highly related retroviruses isolated from gallinaceousbirds. On the basis of sequence comparison and overall genome organization, these viruses are more similar tothe mammalian type C retroviruses than to the avian sarcoma/leukemia viruses. The envelope of a member ofthe REV family, spleen necrosis virus (SNV), is about 50%o identical in amino acid sequence to the envelope ofthe type D simian retroviruses. Although SNV does not productively infect primate or murine cells, thereceptor for SNY is present on a variety of human and murine cells. Moreover, interference assays show thatthe receptor for SNY is the same as the receptor for the type D simian retroviruses. We propose that adaptationof a mammalian type C virus to an avian host provided the REV progenitor.

Spleen necrosis virus (SNV) is a type C retrovirus isolatedfrom ducks that is capable of killing ducklings within a weekof inoculation (35) and inducing immunodeficiency in olderanimals (38). The virus is cytopathic to cultured chickenembryo fibroblasts and can productively infect a canine cellline (32, 33). Phylogenetically, SNV is grouped among theavian reticuloendotheliosis virus (REV) subspecies, whichalso includes the defective, transforming virus REV-T andits replication-competent helper, REV-A (38). SNV andREV-A share over 90% nucleic acid sequence homologyacross their genomes (24) and use a common receptor (8, 16).The predicted amino acid sequence of the envelope gene

of REV-A has been reported to have about 42% amino acididentity with the env gene of the simian type D retrovirusesand the type C baboon endogenous virus (BAEV) (18, 31).Assays based on infection interference demonstrate that thesimian type D retroviruses, BAEV, and the endogenous typeC feline virus RD114 use a common receptor (29). However,the gag and pol genes of the REV family are distinct fromthose of simian type D retroviruses and instead share ho-mology with those of mammalian type C retroviruses (2, 3,24, 37). Despite the similarities in primary sequence ofmammalian retroviruses and REV, REV cannot produc-tively infect cells of murine or primate origin (14, 19).

Here, we show that the REV receptor is present on bothmurine and human cell lines. These cells can be infected withretrovirus vectors that are packaged with REV Gag, Pol, andEnv proteins. While this work was in progress, Koo et al.also reported that SNV can enter human as well as simiancells (19). Furthermore, sequence analysis of SNV env

reveals a greater predicted amino acid sequence homologywith the simian type D retroviruses and BAEV than previouscomparisons of REV-A to these simian retroviruses. Ac-cordingly, preinfection of cells susceptible to REV infectionwith a simian type D virus will inhibit subsequent infectionof REV-A by three orders of magnitude. This interferencesuggests that REV-A and SNV use the same receptor as thesimian type D retroviruses and BAEV. On the basis ofsequence similarities and utilization of a common receptor,

* Corresponding author.

we propose that the progenitor of the avian REVs originatedthrough the adaptive radiation of a BAEV-like mammalianretrovirus.

MATERIALS AND METHODS

Nomenclature. Plasmids have a small "p" before theirname, while virus stocks derived from those plasmids do not(e.g., pLXSH and LXSH). Virus stocks are designated bythe name of the plasmid used to derive the genome followedin parentheses by the type of envelope protein on the virion[e.g., LXSH(REV) is a retrovirus vector derived from trans-fection of pLXSH into a REV-packaging line]. Infected cellsare designated by the name of the cell line followed by thename of the virus (e.g., D17/SNV).

Plasmids. pLXSH is a murine leukemia virus (MuLV)-based retroviral vector identical to pLXSN (22) except thatit contains the hygromycin B phosphotransferase (hph) gene,which confers resistance to the drug hygromycin B ineukaryotic cells, in place of the neo gene of pLXSN.pSNV-hygro is an SNV-based vector derived frompJD216neo (11). It contains the hph gene driven by the SNVLTR. pPB1l1 and pSW253 are complete proviral clones ofSNV and REV-A, respectively (1, 5). pSHRM-15 encodes a

complete proviral clone of SRV-3 (25).Sequencing and analysis. The SNV env open reading frame

and flanking sequences of pPB1O1 were cloned into pBlue-scriptKS+ (Stratagene Co., La Jolla, Calif.) and pTZ18R(Pharmacia). Subclones were generated by making nestedunidirectional deletions commencing from the 5' portion ofthe env gene (17). Single-stranded DNA was produced byusing M13K07 helper phage and was the substrate for Sangerdideoxy sequencing (27). Nucleotide oligomers correspond-ing to phage T3 and T7 RNA polymerase promoters andoligomers based on the REV-A env sequence (37) were usedas primers in sequencing.Sequence analyses were performed with the software

package of the University of Wisconsin Genetics ComputerGroup (9), using the BestFit algorithm (28) for alignment ofamino acid sequence. In all alignments, gap weight was 5.0and gap length weight was 0.1.

Cells. D17 cells are derived from a dog osteosarcoma, NIH

3026

Vol. 66, No. 5

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 06

Jan

uary

202

2 by

125

.177

.140

.13.

SNV RECEPTOR 3027

3T3 tk- is a mouse fibroblast line, and HeLa cells are

derived from a human cervical carcinoma. B78 cells, L aprt-tk- cells, C3H 1OT1/2, and C3H C2C12 cells are of murineorigin. Human primary fibroblasts were a gift of MattThayer. CEMx174 cells are a T/B-cell human hybrid line,and H9 is a human T-cell line. PA317 cells are a packagingline that will package viral genomes in an MuLV virion withan amphotropic envelope (21). D17.2G cells are a packagingline that will package viral genomes in an REV-A virion withan REV-A envelope (12). Unless otherwise indicated, all celllines were grown in Dulbecco's modified Eagle's medium(DMEM) with 10% calf serum. COS-7 cells were grown inDMEM with 10% fetal bovine serum. CEMx174 cells and H9cells were grown in RPMI medium with 10% fetal bovineserum.

Cells chronically infected with SNV were derived bytransfecting D17 cells with pPB1l1. Cell supernatant was

harvested 1 week posttransfection and used to infect freshD17 cells in the presence of 10 ,ug of Polybrene per ml. Thesurviving cells were passaged, and cell supernatants re-mained positive for reverse transcriptase activity over theduration of all experiments (data not shown). SNV reversetranscriptase assays were done in the presence of 0.6 mMMnCl2.

Simian retrovirus serotype 3 (SRV-3) was obtained bytransfecting COS-7 cells with pSHRM-15. Virus producedtransiently was used to infect D17, HeLa, CV-1, CEMx174,and H9 cells. Chronic infection was established only inCEMx174 cells (data not shown). These are calledCEMx174/SRV-3 cells. In order to establish a population ofD17 cells uniformly infected with SRV-3, 5 x 10 D17 cellswere trypsinized and then mixed with 2 x 106 CEMx174/SRV-3 cells and replated in a 100-mm-diameter plate con-taining DMEM plus 10% fetal calf serum and 10 ,ug ofPolybrene per ml. After 2 days, the nonadherent CEMx174/SRV-3 cells were removed and the monolayer was washedthree times with phosphate-buffered saline. These cells,called D17/SRV-3, were passaged by trypsinization at leasttwice more before use. In addition, RNA derived fromD17/SRV-3 cells does not contain a human T4-cell-specificmessage, CD4, which is expressed in CEMx174 andCEMx174/SRV-3 cells (data not shown).COS-7 cells were transfected with DEAE-dextran and

chloroquine as described previously (15). All other transfec-tions were done by the method of Chen and Okayama (4).Total cellular RNA was extracted by the acid guanidiniumthiocyanate-phenol-chloroform protocol (6). RNA blottingand analysis was done by standard techniques (26).

Generation of viral stocks. Cell lines producing virallypackaged vector were obtained by transfection of REV-A-or amphotropic MuLV (A-MuLV)-packaging cells lines(D17.2G and PA317, respectively) with pLXSH followed byselection for hygromycin B resistance. Virus was harvestedfrom packaging cells which were stably transfected with thevector. Harvested viral supernatant was centrifuged to re-move infected cells and stored at -80°C in small aliquots.Vector pseudotypes obtained include LXSH(REV) andLXSH(A-MuLV).Another set of REV-A-packaged vectors was obtained by

cotransfection of D17 cells with the SNV-based vectorpSNV-hygro and pSW253. Viral supernatant was gatheredon successive days after the appearance of Mn2+-dependentreverse transcriptase activity in the cell supernatant. Thisvirus stock is designated REV-hygro.

Infections and virus titering. Cells were plated at 1 x 105 to2 x 105 cells per well on a 12-well plate (22-mm diameter).

TABLE 1. Prevalence of SNV receptor among mammalian cells

Hygromycin-resistant CellCell linea colonies transduced .i.

by LXSH(REV) (%)b origi

D17 10fc CanineHeLa 12 HumanHFFd 59 HumanL aprt- tk- 4 MurineB78 5 MurineC2C12 (C3H) 5 Murine1OT1/2 (C3H) 67 MurineNIH 3T3 tk- <0.02 Murine

a Cell lines are described in Materials and Methods.b Percentages are calculated from the number of drug-resistant colonies

obtained from the particular cell line relative to the number of coloniesobtained from an infection of D17 cells. Numbers from infections wereobtained from representative experiments in which infections were alwayscarried out parallel to infections of D17 cells. Infections were carried out withlimiting dilutions of the retrovirus vector so that the number of resistantcolonies is proportional to virus input on D17 cells.

c Virus titers obtained with D17 cells in the individual experiments variedbetween 3.9 x 104 and 8.0 x 105 hygromycin transforming units per ml.

d HFF, human primary fibroblasts.

The next day, wells were generally 50% confluent. Mediumwas removed, and 300 ,ul of serial virus dilutions containingmedium with serum and 10 ,ug of Polybrene per ml wasadded to each well. After 1 h, the virus inoculum wasremoved and 1 ml of medium was added to each well.Approximately 24 h postinfection, medium was replacedwith a selective medium. Selective media were changedevery 2 to 3 days until the mock-infected cells died (usually10 to 14 days postinfection), and virus titers were calculated.

Nucleotide sequence accession number. The 1,704-nucleo-tide base sequence of the SNV Env open reading frame hasbeen deposited in the GenBank data base under accessionnumber M87666.

RESULTS

REV receptor appears on many mammalian cell types.Because SNV and REV-A do not productively infect humanor murine cells, it was necessary to use a biological assaywhich is independent of virus production to detect thepresence of the REV receptor on these cell types. Tocircumvent difficulties in the expression of SNV-based vec-tors in certain mammalian cell types (14), virions wereengineered that contained REV-A Gag, Pol, and Env pro-teins and that have packaged the genome of a defectivemurine retrovirus instead of an SNV provirus. The viralgenome packaged in this system is the MuLV-based retro-virus vector LXSH. The virus stock, LXSH(REV), washarvested from the REV-A-packaging line D17.2G (12),which is stably transfected with pLXSH. It has previouslybeen shown that REV proteins can package MuLV-derivedvectors 50% as efficiently as SNV-derived vectors in theabsence of a competing SNV vector (13).The presence of the REV receptor on human and mouse

cells was determined by the ability of LXSH(REV) to conferdrug resistance after infection. The virus titer, measured byinfection of each human and mouse cell line, was comparedwith the titer measured by parallel infection of a canine cellline (D17) that is permissive for REV infection (Table 1).Surprisingly, we found that LXSH(REV) can efficientlyinfect human cells. There were 12% as many drug-resistantHeLa colonies as there were drug-resistant D17 colonies

VOL. 66, 1992

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 06

Jan

uary

202

2 by

125

.177

.140

.13.

3028 KEWALRAMANI ET AL.

SNV 1 MDCLTLR5AFGKVDQAKILILLVA1I I

SRV-3 1 MNFIWSLVILSQISQvQA....

100

83 .

190

181

283

281

9FGTTAEGYP

1111 11 III 11

?.LQQLWELPCDCSGGYVSSIPTYYTYSLDCGGSThYLTYGSGGSWSWGGFKQWCVFKPill 11 11 1111111 ill 11

..GFGDPREAAIQQKHGKPCDCAGGYVSSPPINSLTTVSCSTHTAY....SVTNSL ........KWQCVSTP

V V1., P IQASCTM.I 1'VGKPVWI PVAPVYVSDGGG

11 III 11111II.TTTHIGSCPCTISYDSVHASCYNHYQQCNIGNTYLTATITGDRTPAIGDGNVPTVLGTSHNLITAGCPN QWCWNSERPSVBISDGGG

PTDMIREESVRPEMLEIIRBSYPSVQYHPLAP PRSVD IVP PQSDI LNATNPKUAKNAWNCMLGIPIPIPT .....I LDNS

I I I 11 illlll 11 III III i 11 1 1 1 1 1PQDRARDIIVNFLHERSLFPELSYYHPILEARKIAElDALDLATVHSLLNASQPSLAEDCWLCLQSGDPVPLALPYNJDTlSNFACLSNHSCP

v E 0o vvLSLPFGCNPPGSIDVSC.YAGEADNRTGIPVGYVHFTNCTSIQEVTNETSQMNTRLCPPPGVVCGNMAYTALPwPwIGcILAsIvPDISIIsGE

11 11 I 11 1iii 11 1111111 III 11 11 III 11LTPPFLVQPFWFTDSNCLYAEYQNNSFDIDVGLASFTNCSSYYNVSTASKPSNS... ICAPNSSVFVCGNNKAYTYIPTNWTGSCVLATLLPDIDIIPGS

382 EPIPLPSIEYTARRHK RAVQFIPLLVGLGISGATLAGGTGLGVSVETYHKLSNQLIEDVQALSGTINDLQDQIDSLAKVVLQNRRGLDLLTAEQG11 11 1 1 * 11 1111 11111 1 1 11111 1 111 111 1111 1 1111111111111111111

378 EPVPIPAIDHFLGKAK RAIQLIPLFVGLGITT!AVSTGAAGLGVSITQYTKLSHQLISDVQAISSTIQDIQDQVDSLAEVVLQNMGLDLLTAEQGV W v

477 GICLAIQEKCCFYANKSGIVRDKIRKQEDLIERKRALYDNPLWSGLNGFLPYLPLLGPLFGLILFLTLGPCIMKTLTRIIHDKIQAVKS 5671111111111111111111111 11 11 1 1 1111 1 11111 111111 1I1 11 1I

473 GICIQEKCCFYANKSGIVRDKIKNL DDIRRRRQLIDNPF TSFEGFLPYVIMPLGPLLCLLLVLSFGPIIFNKLHTFIKHQIESIQA 563

FIG. 1. Alignment of the predicted amino acids of the SNV and SRV-3 envelope proteins. The nucleic acid sequence of SNV wastranslated to a one-letter amino acid code and aligned with the amino acid sequence of SRV-3 as described in Materials and Methods. Dotsalong the sequences represent gaps in the alignment. Vertical lines represent identical amino acids. SNV paired cysteines are denoted by /.SNV glycosylation sites which nearly align with SRV-3 (within 3 amino acids) are denoted by X, and the site which does not align is denotedby 0. The cleavage between the SU and the TM proteins is marked with an asterisk. The immunosuppressive peptide is underlined.

(Table 1). Moreover, the primary human skin fibroblastculture gave 59% as many colonies as D17 cells. Thisindicates that despite the block to productive infection ofSNV in human cells, the receptor for SNV must exist onhuman cells. In addition, because the colonies can be grownfor multiple generations, it is unlikely that there is a block toproviral integration.

In agreement with previous reports that indicated thatNIH 3T3 cells could not be infected with SNV and that oneof the blocks to infection may be at entry (14), we could notdetect infection of NIH 3T3 cells with our pseudotypedvector (Table 1). However, a survey of four other mousecell lines showed that each could be infected. One mousecell line, 1OT1/2, could be infected very efficiently byLXSH(REV) and gave 67% as many hygromycin-resistantcolonies as did D17 cells. The difference in sensitivity toinfection is not mouse strain specific because 10T1/2, C2C12,and L cells are derived from C3H mice. Supernatant ob-tained from LXSH-transformed murine clones of B78 and1OT1/2 cells did not contain reverse transcriptase activityand was not able to transfer drug resistance to naive B78 or1OT1/2 cells (data not shown). This indicates that infection ofmouse cells was not due to contamination with an ampho-tropic or ecotropic helper. Moreover, infection of NIH 3T3cells with the same stocks used to infect B78, 1OT1/2, and theother mouse cell lines yielded no drug-resistant colonies.Thus, the receptors for the avian retroviruses REV-A andSNV are displayed on some cells of murine and humanorigin.SNV Env shares homology with Env of BAEV and simian

type D retroviruses. Of those proteins which are common toall retroviruses, the envelope (Env) protein is the least

conserved (10, 20). Within the simian type D retrovirusreceptor interference group, the conservation of amino acididentity between the Env proteins of the different membersranges from 48 to 87% (references 18, 23, and 34 and data notshown). We sequenced the env gene of an infectious molec-ular clone of SNV and determined the degree of similarity itshared with the Env proteins of other retroviruses that infectmammalian cells. We found that the predicted amino acidsequence of SNV Env is 49% identical to that of a prototypicsimian type D virus, SRV-3 (Fig. 1). When conservativeamino acid changes are included in the alignment, there is65% similarity between the two sequences. SNV Env is evenmore highly conserved with the squirrel monkey retrovirus(55% amino acid identity) and the type C BAEV (54%identity) (Table 2). More importantly, however, the posi-tions and occurrences of cysteines within the SNV Env areconserved among the Env proteins of the members of thesimian type D virus group as well as the Env of the type CBAEV (Fig. 1 and Table 2). This suggests that the tertiarystructures of the Env proteins determined by disulfide bondsare probably similar among these simian viruses. In addition,there is a very strong conservation of the potential N-linkedglycosylation sites between SNV Env and the Env proteinsof the viruses in the simian type D retrovirus interferencegroup.

Within Env, the transmembrane (TM) portion of the Envprecursor is more conserved than the surface (SU) portionbetween SNV and the envelope of BAEV and the simiantype D viruses (Fig. 1 and data not shown). The specificregions of high identity within the TM portion includeresidues designated the "immunosuppressive peptide,"which appear in most mammalian type C and type D

KIIPSVQGQPGPCPSECLQIATQ.MHSTCYEKTQECTLLGKTYFTAILQKTKlGSYEDG...

J. VIROL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 06

Jan

uary

202

2 by

125

.177

.140

.13.

SNV RECEPTOR 3029

TABLE 2. Similarity of type D and type C retroviral envelopesto SNV envelope

No. of putative N-

Virus' iAnmo acid No. of aligned linked glycosylation(%)b cysteines/total no.c sites conserved/

SRV-1 49.5 19/23 7/8SRV-2 50.2 19/23 7/8SRV-3 49.3 21/23 7/8SMRV 55.2 20/23 7/8BAEV 53.6 20/23 6/8GALV 29.8 9/23 3/8Mo-MuLV 28.0 8/23 3/8

a Virus sequence information was obtained from the GenBank data base(Los Alamos National Laboratory). SRV-1, -2, and -3 (simian retrovirusserotypes 1, 2, and 3) and SMRV (squirrel monkey retrovirus) are type Dretroviruses, whereas BAEV, GALV (gibbon ape leukemia virus), andMo-MuLV (Moloney murine leukemia virus) are type C retroviruses.

b Alignments were performed by using the BestFit algorithm (28) in theUniversity of Wisconsin Genetics Computer Group software package (9). Gapweights = 5.00; gap length weights = 0.1.

c These ratios reflect the number of cysteins which align against the totalnumber of SNV cysteines under the BestFit algorithm. The mature form ofSNV Env glycoprotein contains a total of 23 residues in the SU and TMportions of the molecule and does not contain the N-terminal secretory leadersequence included in Fig. 1.

d Putative N-linked glycosylation sites are those tripeptides of the sequenceN-X-S or N-X-T, where X cannot be a proline residue. Sites are consideredaligned when they occur within 3 residues under the amino acid alignment.

retroviruses, and the regions which flank the immunosup-pressive peptide (7, 31). In contrast, SNV displays only 28%amino acid identity with Moloney murine leukemia virus(Mo-MuLV) and 30% with another type C primate retrovi-rus, gibbon ape leukemia virus. The limited homology ob-served with these type C viruses appears outside of the SUportion and is due to the highly conserved immunosuppres-sive peptide in the TM portions of the viruses (reference 31and data not shown). The unusual degree of similarity ofSNV Env with Env of the type D retroviruses and of BAEV,in fact, predicts a phylogeny that is distinct from pol-basedtrees which place SNV at a closer evolutionary distance toMuLVs than to the simian type D viruses or BAEV (10, 20).

Interference of REV infection by preinfection of cells withSRV-3. Given the similarity in Env between SNV, BAEV,and the simian type D viruses, we wished to determinewhether these viruses use the same receptor. Interferenceassays have been used to determine whether or not tworetroviruses use the same receptor (29, 36). Interference isthought to occur by the association of the endogenouslyexpressed Env with the cellular receptor, thus reducing theamount of receptor that may function on the cell surface(36). Preinfection of cells with SNV or REV-A will protectcells from superinfection by a REV-packaged vector throughthe endogenous production of retroviral Env (8, 16). Wetherefore sought to determine whether preinfection of cellswith SRV-3, a simian type D virus, would interfere withsubsequent infection of those cells by REV-A.D17 cells were used as a target for this experiment because

they are readily infectable by REV-A and SNV. D17 cellschronically infected with SNV served as the positive controlfor interference of infection by a REV-packaged virion (8,16). On the other hand, we could not produce a population ofD17 cells that were chronically infected with SRV-3 bydirect infection because the virus could not spread in theculture (data not shown). SRV-3 could readily infect andreplicate in the human lymphoid cell line CEMx174. There-

Hygror colonies/ml

D17 D17/SRV-3 D17/SNV

REV-Hygro

hyg 1.9 x 10 4 1.5 x 10 I 1.5x 101,

LXSH(A-MuLV)

EJ=SV4@ hyg C==E 6.8 x 104 5.9 x 104 6.3 x 104FIG. 2. Interference of REV infection by preinfection of cells

with SRV-3. Stocks of REV-hygro with REV-A helper virus andstocks of helper-free LXSH(A-MuLV) produced from an amphotro-pic MuLV-packaging line were prepared as described in Materialsand Methods. Dilutions of each virus stock were used to infect D17cells, D17 cells preinfected with SRV-3, and D17 cells preinfectedwith SNV. The number of hygromycin-resistant colonies per ml ofvirus stock is shown. A titer of 1.5 x 101 hygromycin-resistant(Hygror) units per ml is estimated from three colonies in a well thatwas infected with 200 ,ul of virus. A representative experiment isshown.

fore, in order to obtain a population of D17 cells that wasuniformly infected with SRV-3, we cocultivated the nonad-herent SRV-3-infected CEMx174 cells (CEMx174/SRV-3cells) with a monolayer of D17 cells to produce D17/SRV-3cells (for details, see Materials and Methods).

In order to demonstrate that infected cells expressed envof either SRV-3 or SNV, RNA from D17 cells, D17/SRV-3cells, and D17/SNV cells was extracted and analyzed. Hy-bridization of a Northern (RNA) blot using a probe fromSRV-3 probe encompassing the 5' end of the env gene showsthat the D17 cells infected with SRV-3 by cocultivation withproducer cells contain two species of RNA corresponding tothe genomic RNA and spliced envelope RNA (data notshown). The ratio of envelope to genome RNA from theinfected D17/SRV-3 cells parallels the ratio of viral RNAfrom productively infected CEMx174/SRV-3 cells. Probing aparallel blot with a REV-specific probe showed SNV ge-nomic and envelope RNA only in the D17 cells chronicallyinfected with SNV (data not shown). These results demon-strate that neither culture was cross-contaminated.These cells were then challenged with a REV-A-packaged

virus, REV-hygro. As shown in Fig. 2, this stock yielded 1.9x 104 hygromycin-resistant colonies per ml of inoculumwhen D17 cells were infected. This same stock yielded only15 hygromycin-resistant colonies per ml when it was used toinfect D17 cells that were chronically infected with SNV.This confirms previous findings that REV-A and SNV use acommon receptor (8, 16).The REV-hygro stock was also used to infect D17/SRV-3

cells. Preinfection of the D17 cells with SRV-3 reduced thenumber of hygromycin-resistant colonies to 15 per ml ofREV-hygro (Fig. 2). This indicates that SRV-3 interfereswith REV infection and therefore implies that both virusesuse the same receptor. In order to demonstrate that thisinterference was specific to vectors packaged with REVproteins, the titer of an amphotropic virus stock containingthe MuLV vector LXSH was determined on D17, D17/SNV,and D17/SRV-3 cells. The titer of LXSH(MuLV) was notsignificantly affected by preinfection of the D17 cells witheither SNV or SRV-3 (Fig. 2). This indicates that theresistance to REV-hygro infection of D17 cells preinfectedwith SNV or SRV-3 is specific to the challenge virus.

VOL. 66, 1992

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 06

Jan

uary

202

2 by

125

.177

.140

.13.

3030 KEWALRAMANI ET AL.

DISCUSSION

The sequence of the env gene of SNV, an avian virus,reveals approximately 50% identity with the predicted aminoacid sequences of both the simian type D retroviruses andBAEV. Cells preinfected with either SRV-3 or SNV aremarkedly less susceptible to superinfection when challengedby vectors packaged with REV proteins. The interference toinfection suggests that these avian and simian viruses use anequivalent receptor. Consistent with this conclusion, wefound that viral pseudotypes displaying a REV Env caninfect human cells. We also find that REV-A pseudotypescan enter some murine cells. Blocks to productive infectionof the virus in these mammalian cells, therefore, must occurat steps in the virus life cycle beyond virus entry.The presence of a REV receptor on murine cells is

somewhat surprising, because the simian type D viruses arerestricted from entering mouse cells (29, 30). On the otherhand, pseudotypes made with the BAEV envelope can infectmurine cells at low efficiency (30). These inconsistencies inreceptor tropism are probably due to subtle differences in theEnv proteins of viruses in this interference group (29). Inaddition to SNV Env and BAEV Env pseudotypes beingable to enter mouse cells, the gag and pol gene products ofthe REVs and type C BAEV are more closely related to oneanother than they are to the type D simian retroviruses (2, 3,24, 36, 37).These data imply a common evolutionary origin for SNV

and BAEV that is distinct from the presumed recombinato-rial origin of the simian type D retroviruses, which possess amouse mammary tumor virus-like Gag and Pol (18, 31).Unlike the REVs, BAEV is an endogenous virus and istherefore transmitted as a stable genetic element. Also,unlike SNV, BAEV displays no apparent pathogenesis to itshost (36). Because SNV is an exogenous and pathogenicretrovirus, we propose that the avian REVs are the result ofan adaptive radiation of a BAEV-like, mammalian retrovirusto birds. Direct transmission of BAEV to the North Ameri-can fowls which host the REV-like viruses seems unlikely,because both viral hosts are geographically isolated and thexenotropic host range of BAEV does not extend to aviancells in vitro (36). Because the variability of BAEV is fixedas an endogenous element, the restriction of BAEV fromavian cells precludes the possibility of a REV being thedirect progenitor of BAEV. This restriction also suggests theexistence of another type C virus, related to BAEV, whichwas the true REV progenitor. Given the potential of retro-viruses to spread among similar host species, retroviraltransmission across animal phyla has not been previouslydocumented to have occurred in recent viral evolution. SNVmay be exceptional in that regard.

ACKNOWLEDGMENTSWe thank Peggy Lee, Paul Lewis, and Maxine Linial for discus-

sions and critical readings of the manuscript. V.N.K. thanks EricDelwart for insightful suggestions. We thank Eric Hunter, A. DustyMiller, Howard Temin, and Matt Thayer for plasmids or cells andJoe Dougherty for communicating results prior to publication.

This work was supported by grants R01 AI30927 (M.E.) and CA22443 (A.T.P.) from the Public Health Service. V.N.K. is supportedby NRSA T32 GM07270 from the National Institute of GeneralMedical Sciences. A.T.P. is a Shaw Scholar. M.E. is a Scholar ofthe American Foundation for AIDS Research.

REFERENCES1. Bandyopadhyay, P. L., and H. M. Temin. 1984. Expression from

an internal AUG codon of herpes simplex thymidine kinase geneinserted in a retrovirus vector. Mol. Cell. Biol. 4:743-748.

2. Barbacid, M., E. Hunter, and S. A. Aaronson. 1979. Avianreticuloendotheliosis viruses: evolutionary linkage with mam-malian type C retroviruses. J. Virol. 30:508-514.

3. Charman, H. P., R. V. Gilden, and R. L. Witter. 1979. Reticu-loendotheliosis virus: detection of immunological relationship tomammalian type C retroviruses. J. Virol. 29:1221-1225.

4. Chen, C., and H. Okayama. 1987. High-efficiency transforma-tion of mammalian cells by plasmid DNA. Mol. Cell. Biol.7:2745-2752.

5. Chen, I. S. Y., T. W. Mak, J. J. O'Rear, and H. M. Temin. 1981.Characterization of reticuloendotheliosis virus strain T DNAand isolation of a novel variant of reticuloendotheliosis virusstrain T by molecular cloning. J. Virol. 40:800-811.

6. Chomczynski, P., and N. Sacchi. 1987. Single-step method ofRNA isolation by acid guanidinium thiocyanate-phenol-chloro-form extraction. Anal. Biochem. 162:156-159.

7. Cianciolo, G. J., T. D. Copeland, S. Oroszlan, and R. Snyder-man. 1985. Inhibition of lymphocyte proliferation by a syntheticpeptide homologous to retroviral envelope proteins. Science230:453-455.

8. Delwart, E. L., and A. T. Panganiban. 1989. Role of reticuloen-dotheliosis virus envelope glycoprotein in superinfection inter-ference. J. Virol. 63:273-280.

9. Devereux, J., P. Haeberli, and 0. Smithies. 1984. A comprehen-sive set of sequence analysis programs for the VAX. NucleicAcids Res. 12:387-395.

10. Doolittle, R. F., D. F. Feng, M. A. McClure, and M. S. Johnson.1990. Retrovirus phylogeny and evolution, p. 1-23. In R.Swanstrom and P. K. Vogt (ed.), Retrovirology. Springer-Verlag, Berlin.

11. Dougherty, J. P., and H. M. Temin. 1986. High mutation rate ofa spleen necrosis virus-based retrovirus vector. Mol. Cell. Biol.6:4387-4395.

12. Dougherty, J. P., and H. M. Temin. 1988. Determination of therate of base-pair substitution and insertion mutations in retro-virus replication. J. Virol. 62:2817-2822.

13. Embretson, J. E., and H. M. Temin. 1987. Lack of competitionresults in efficient packaging of heterologous murine retroviralRNAs and reticuloendotheliosis virus encapsidation-minusRNAs by the reticuloendotheliosis virus helper cell line. J.Virol. 61:2675-2683.

14. Embretson, J. E., and H. M. Temin. 1987. Transcription from aspleen necrosis virus 5' long terminal repeat is suppressed inmouse cells. J. Virol. 61:3454-3462.

15. Emerman, M., R. Vazeux, and K. Peden. 1989. The rev geneproduct of the human immunodeficiency virus affects envelope-specific RNA localization. Cell 57:1155-1165.

16. Federspiel, M. J., L. B. Crittenden, and S. H. Hughes. 1989.Expression of avian reticuloendotheliosis virus envelope con-fers host resistance. Virology 173:167-177.

17. Henikoff, S. 1987. Unidirectional digestion with exonuclease IIIin DNA sequence analysis. Methods Enzymol. 155:156-165.

18. Kato, S., K. Matsuo, N. Nishimura, N. Takahashi, and T.Takano. 1987. The entire nucleotide sequence of baboon endog-enous virus DNA: a chimeric genome structure of murine typeC and simian type D retroviruses. Jpn. J. Genet. 62:127-137.

19. Koo, H.-M., A. M. C. Brown, Y. Ron, and J. P. Dougherty.1991. Spleen necrosis virus, an avian retrovirus, can infectprimate cells. J. Virol. 65:4769-4776.

20. Lewe, G., and R. M. Flugel. 1990. Comparative analysis of theretroviralpol and env protein sequences reveal different evolu-tionary trees. Virus Genes 3:195-204.

21. Miller, A. D., and C. Buttimore. 1986. Redesign of retroviruspackaging cell lines to avoid recombination leading to helpervirus production. Mol. Cell. Biol. 6:2895-2902.

22. Miller, A. D., and G. J. Rosman. 1989. Improved retroviralvectors for gene transfer and expression. BioTechniques 7:980-990.

23. Oda, T., S. Ikeda, S. Watanabe, M. Hatsushika, K. Akiyama,and F. Mitsunobu. 1988. Molecular cloning, complete nucleotidesequence, and gene structure of the provirus genome of aretrovirus produced in a human lymphoblastoid cell line. Virol-ogy 167:468-476.

J. VIROL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 06

Jan

uary

202

2 by

125

.177

.140

.13.

SNV RECEPTOR 3031

24. Panganiban, A. T. Unpublished data.25. Rhee, S. S., and E. Hunter. 1990. Preassembled capsids of type

D retroviruses contain a signal sufficient for targetting specifi-cally to the plasma membrane. J. Virol. 64:3844-3852.

26. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecularcloning: a laboratory manual, 2nd ed. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y.

27. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-

ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

28. Smith, T. F., and M. S. Waterman. 1981. Identification ofcommon molecular subsequences. J. Mol. Biol. 147:195-197.

29. Sommerfelt, M. A., and R. A. Weiss. 1990. Receptor interfer-ence groups of 20 retroviruses plating on human cells. Virology176:58-69.

30. Sommerfelt, M. A., B. P. Williams, A. McKnight, P. N. Good-fellow, and R. A. Weiss. 1990. Localization of the receptor genefor type D simian retroviruses on human chromosome 19. J.Virol. 64:6214-6220.

31. Sonigo, P., C. Barker, E. Hunter, and S. Wain-Hobson. 1986.Nucleotide sequence of Mason-Pfizer monkey virus: an immu-nosuppressive D-type retrovirus. Cell 45:375-385.

32. Temin, H. M., and V. K. Kassner. 1974. Replication of reticu-

loendotheliosis viruses in cell culture: acute infections. J. Virol.13:291-297.

33. Temin, H. M., and V. K. Kassner. 1975. Replication of reticu-loendotheliosis virus in cell culture: chronic infection. J. Gen.Virol. 27:267-274.

34. Thayer, R. M., M. D. Power, M. L. Bryant, M. B. Gardner,P. J. Barr, and P. A. Luciw. 1987. Sequence relationships oftype D retroviruses which cause simian acquired immunodefi-ciency syndrome. Virology 157:317-329.

35. Trager, W. 1959. A new virus of ducks interfering with devel-opment of malaria parasite (Plasmodium lophurae). Proc. Soc.Exp. Biol. Med. 101:578-582.

36. Weiss, R. A., N. M. Teich, H. A. Varmus, and J. Coffin. 1985.RNA tumor viruses. Cold Spring Harbor Laboratory, ColdSpring Harbor, N.Y.

37. Wilhelmsen, K. C., K. Eggleton, and H. M. Temin. 1984.Nucleic acid sequences of the oncogene v-rel in the reticuloen-dotheliosis virus strain T and its cellular homolog, the proto-oncogene c-rel. J. Virol. 52:172-182.

38. Witter, R. L. 1984. Reticuloendotheliosis, p. 406-417. In M. S.Hofstrad, H. J. Barnes, B. W. Calnek, W. M. Reid, andH. W. J. Yoder (ed.), Diseases of poultry. Iowa State Univer-sity Press, Ames.

VOL. 66, 1992

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

vi o

n 06

Jan

uary

202

2 by

125

.177

.140

.13.