a chimeric envelope gene which encodes the external domain. 16116 ... cells were infected with 50-100 plaque-forming units of TK- vaccinia virus, in the ... methionine-free medium and incubated in methionine-free medium for 30 min.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 262, No. 33, Issue of November 25, pp. 16116-16121,1987 Printed in U.S.A .
Effects of Deletion of the Cytoplasmic Domain upon Surface Expression and Membrane Stability of a Viral Envelope Glycoprotein* (Received for publication, May 28,
D. R. KilpatrickS, R. V. Srinivas, E.B. Stephens, andR. W.Compans From the Department of Mkrobiology, University of Alabama at Birmingham, Birmingham, Alabama 35294
The envelope protein (gp52) of Friend spleen focusforming virus (F-SFFV) is defective in its intracellular transport and accumulates in the rough endoplasmic reticulum of F-SFFV-infected cells. This defect in transport has been attributed to the lack of a cytoplasmic domain, and possible loss of signals required for transport to the cell surface. The mature form of gp52, designated gp65, is also reported to be secreted from SFFV-infected cells. To determine the specific changes in the envelope protein which may lead to its lack of transport and to its lackof stability in associating with membranes, the 3’ end of the F-SFFV envelope gene, which encodes the transmembrane domain, was inserted in place of the normal 3’end of the Friend murine leukemia virus genome. This chimeric envelope gene was expressed using the vaccinia virus expression system. The chimeric gp7O/p15E glycoproas tein molecule lacks thecytoplasmic tail residues and a consequence is about 3300 daltons smaller. The chimeric PrEnv molecule was found to be cleaved efficiently as indicated by pulse-chase experiments. Immunofluorescence studiesdemonstrate that the chimeric molecule is efficiently transported to the surface of cells, unlike the SFFV gp52 glycoprotein. The chimeric molecule was found to be unstable in its membrane association and is released into the culture medium. These results indicate that the changes in the membrane spanning region and the lack of a cytoplasmic tail do not determine the defective transport of gp52, but may determine the stabilityof its association with membranes.
gene mink cell focus-forming viruses (Amanuma et al., 1983; 1983). The carboxylWolff et al., 1983; ClarkandMak, terminal region of the SFFVenu gene is closely related to the 3‘ half of MuLV p15E(Wolff et al., 1983). Gp52 is the primary translation product of the SFFV envelope gene, whereas the enu gene of F-MuLV codesfora precursor glycoprotein (PrEnv) which is proteolytically cleaved into two proteins: an amino-terminal glycosylated protein designated gp70 and a carboxyl-terminal, nonglycosylated protein, designated p15E, which spans the membrane bilayer(Dickson et al., 1984). During virus maturation, p15E is further processed by a virusencoded protease into p12E and a small peptidedesignated as the R peptide (Sutcliffe et al., 1980). The SFFV envelope gene contains a large deletion of 585 bases at position 1221 that removes the carboxyl-terminal of portion of gp70 sequences and the amino-terminal portion the p15Esequences, resulting ina fusion protein. In addition, the sequences coding for the membrane spanning region of gp52, as compared to similar domainsin MuLV p15E, reveal the insertion of a six-base tandem repeat at position 1408 which adds two hydrophobic leucine residues. There is alsoa single base insertion at position 1426 that causes a frameshift mutation and results in premature termination of the molecule 34 codons prior to the termination codon in MuLV p15E (Amanuma et al., 1983; Wolff et al., 1983; Clark and Mak, 1983). These changes result in a hydrophobic carboxyl terminus which appears buried in the membrane with no portion of the molecule exposed on the cytoplasmic side of the membrane (Srinivas and Compans,1983a). Unlike the MuLV envelope proteins, gp52 is defective in transport to the cell surface; only 3-5% is expressed on the surface of cells in a processed form designated gp65 (Ruta The Friend spleen focus-forming virus has been identified and Kabat,1980). However, once gp65 does getto the surface, it is reported to be secreted from cells (Pinter and Honnen, as a recombinant MuLV’ containingsubstitutionsinthe envelope gene as well as deletions in the enu, gag, and pol 1985). The proteins specified by other known retroviral onregions of the genome (Amanuma et al., 1983). Genetic studies cogenes appear to possess a substantial diversity as to their of SFFV have shown that this uniqueenu gene is required for subcellular location and biological properties. The glycoproinitiation of acute erythroleukemia by the virus (Linemeyer tein of F-SFFV has not been shown to possess kinase activiet al., 1982; Ruta et al., 1983; Machida et al., 1984). Recent ties, as is thecase with several other retroviral transforming studies indicate that the amino-terminal portion of gp52 is proteins. SFFV also neither contains cellular oncogenes nor closely related to the amino-terminalregion of the envelope has the ability to transform fibroblasts in culture, unlike the replication defective acute transforming viruses. The expres*This study was supported by Grants CA 40440 and CA 18611 sion of gp52 is necessary for the disease process. The analysis from the National Cancer Insfitutc. The costs of publication of this of SFFV mutants has previously shownthat mutations within article were defrayed in part by the payment of page charges. This the envelope gene eliminated the biological activity of the article must therefore be hereby marked “aduertisement” in accordgenome, while mutations outside the envelope gene had no ance with 18 U.S.C. Section 1734 solely to indicate this fact. 4 Supported by National Institutes of Health Postdoctoral Fellow- affect on the leukemogenicity of the virus (Linemeyer et al., ship Award CA 07841. 1982; Machida et al., 1984). It is still unknownhow the SFFV The abbreviations used are: MuLV, murine leukemia virus; SFFV, enu gene products function in the induction of oncogenesis. spleen focus-formingvirus; MDCK, Madin-Darby canine kidney cells; We wished to determine which changes in the SFFV enveHEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid EGTA, lope gene are responsiblefor the defective transportand [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; PBS,phosphatebuffered saline; SDS, sodium dodecyl sulfate; h.p.i., hours post- intracellular accumulationof gp52. Therefore, we constructed a chimeric envelope gene which encodes the external domain infection.
Membrane Stability of a Chimeric gp70/p15E Molecule derived from F-MuLV, and the 3' end of the p15E region from F-SFFV. This constructionincorporates the six-base and single-base insertions from SFFV into the F-MuLV envelope gene, and results in theloss of the cytoplasmic domain for theglycoprotein. The effect of these two insertions on the transport and processing of MuLV gp70 was analyzed using recombinants of vaccinia viruses which carried either thewild type gp70/p15E or the chimeric gp70/p15E. Our studies show that the loss of the cytoplasmic tail in the chimeric glycoprotein did not block its intracellular transport. The chimeric gp70/p15E was transported only to the basolateral surface of MDCK cells, like the wild type glycoprotein. The rates of cleavage of the wild type PrEnv and the chimeric PrEnv molecules were compared, and the stability of the chimeric and wild type gp70/p15E on cell membranes after transport to the cell surface was determined. MATERIALS AND METHODS
virus stocks were prepared in CV-1 cells. Immumprecipitation of Radiolabeled Proteins-TK-143 cells or CV-1 cells were infected with vaccinia virus recombinants, which contained either the w t gp70/p15E or the chimeric gp70/p15E, at an multiplicity of infection of 10. At 5.5h.p.i. cells were washed with methionine-free medium and incubated in methionine-free medium for 30 min. At 6 h.p.i. cells were labeled with [35S]metbionine(100 pCi/ml) in methionine-free medium for 15 min. Cells were then washed 3 times with medium and chased with medium containing excess cold methionine, or processed immediately for precipitation. At each chase period, the supernatant was removed and clarified at 13,000 X g for 10 min and collected on ice until the end of the experiment a t which time 2 pl of antiserum prepared against Rauscher murine leukemia virus (R-MuLV) gp70was added and incubated overnight at 4 "C with constant mixing. For immune precipitation of cell-associated polypeptides, at each time point the cell monolayers were washed 3X with cold phosphate-buffered saline (PBS), lysed with 0.25 ml of lysis buffer (50 mM Tris-HC1 (pH 7.5), 0.15 M NaCl, 1%Triton X-100, 0.1% SDS, 20 mM EDTA), collected, and after mixing well, centrifuged at 13,000 X g. The cell lysates were incubated with 2 p1 of antiserum against gp70 overnight at 4 'C with constant mixing. Radiolabeled polypeptides were immunoprecipitated with protein A-agarose (Bethesda Research Laboratories). The precipitates were washed four times with cold lysis buffer, resuspended in sample buffer, boiled for 5 min, and analyzed on 10% SDS-polyacrylamide gels (Laemmli, 1970). The protein bands were visualized by fluorography. Immunofluorescence Studies-Either TK-143 or MDCK cells were grown on glass coverslips and infected with vaccinia virus ora vaccinia virus recombinant at an multiplicity of infection of 10. For analysis of internal fluorescence, a t 5.5 h.p.i. cells were washed with PBS, 1%bovine serum albumin and fixed with 95% ethanol, 5% acetic acid for 20 min at -20 "C. Cells were then reacted with antiserum to gp70 for 20 min followedby fluorescein-conjugated rabbit anti-goat IgG. For analysis of surface fluorescence, cells were reacted with antiserum followedby fluorescein anti-goat IgG and then fixed for 15 min with 1%formalin in PBS. For detection of antigen on basolateral cell surfaces, cells were treated with 30 mM EGTA for 20-30 min before staining. The cells were washed with PBS, 1%bovine serum albumin, mounted, and observed for fluorescence using a Nikon Optiphot microscope equipped with a modified B2 cube.
Cells and Virus-CV-1 and MDCK cells were maintained in Dulbecco's modified Eagle's mediumsupplemented with 10% fetal bovine serum. Human TK-143 cells were maintained in the above medium supplemented with 25pg/ml of 5-bromo-2-deoxyuridine. Vaccinia virus stocks were prepared using CV-1 cells and titered in cell lines used for expression studies. Construction of Recombinant VacciniaVirus-The plasmid p231 which contains the genome of Friend murine leukemia virus (FMuLV) was kindly provided by A. Oliff (Memorial Sloan-Kettering Cancer Center, New York, NY). The plasmid p4 which contains the genome of Friend spleen focus-forming virus (F-SFFV) was kindly provided by D. Linemeyer (Merck Institute for Therapeutic Research, Rahway, NJ). The entire envelope gene of p231 was subcloned by digesting p231 with XbaI-KpnI followed by gel isolation of the 2.4kilobase pair fragment. This fragment was ligated to the plasmid vector pGem3 which was digested with XbaI-KpnI. The enu gene of F-SFFV was subcloned by digesting p4 with HindIII-KpnI and gel isolating with 1.8-kilobase pair fragment. This fragment was ligated to theplasmid vector pUC19 which bad been cut with HindIII-KpnI. The ligated DNAs were then used to transform HBlOl cells and the resulting transformants were screened with either an MuLV gene probe or an SFFV gene probe (Rigby et al., 1977) by colony hybridiRESULTS zation. To construct the chimeric envelope gene, the enu gene clones were Construction of a Chimeric (F-MuLVISFFV) Envelope digested with AsuII-KpnI followed by separation of the DNA frag- Gene-To construct a chimeric env gene whichcontained the ments on a low melting temperature agarose gel. The large AsuIIecotropic domain of F-MuLV and the membrane spanning KpnI fragment from the MuLV enu gene digestion was cut out of the gel and ligated to thesmall AsuII-KpnI gel isolated fragment of SFFV. region of SFFV, the AsuII-KpnI fragment of F-MuLV, which The ligated DNA wasused to transform HBlOl cells and theresulting contains the membrane spanning region and the cytoplasmic transformants were screened with an SFFV gene probe. The insertion domain of p15E, was replaced with the AsuII-KpnI fragment of the SFFV membrane spanning region into the F-MuLV enu gene from SFFV. The coding regions of both MuLV and SFFV was confirmed by restriction enzyme analysis. This chimeric gene which are contained within the AsuII-KpnI restriction fragwas then inserted into the plasmid vector pSCll by first cutting the ments are shown in Fig. 1. The amino acid compositions of chimeric enu plasmid with XbaI-Asp718, followed by treatment with these two regions are very similar. The main differences are Klenow DNA polymerase I to obtain blunt ends. This DNAwas ligated to pSCll cleaved with S m I . The enu gene of SFFV was the six-base insertion in SFFV which adds 2 leucine residues similarly inserted into SmaI-cleaved pSCll by cutting the P4 clone HuLV PheGluGlyLeuPheAsnArgSerProTrpPheThrThrLeulleSerThrlle which contains the SFFV genome with BamHI and Asp718,gel I t serser I I c t I I I I t I Ala* isolating this enu geneDNA fragment, and treating with Klenow SFFV DNA polymerase I to obtain blunt ends. HBlOl cells were transformed and screened using an SFFV gene probe. The orientation of MuLV PletGlyProLeullelleLeuLeu - - LeulleLeuLeuPheGlqProCqs the genes in pSCll was determined using restriction enzyme analysis. t ser t t I t L ~ * ~t L IleTrpThrLeu ~ ~ SFFV To generate vaccinia virus recombinant, TK-143 cells were infected with vaccinia virus (strain IHD-J) at an multiplicity of infection of nuLV lleLeuAsnArgLeuVelGlnPheValLysAspArglleSerValValGlnAle 0.05. At 2 h after infection, cells were transfected with a calcium SFFV T y r S e r phosphate precipitate of 10 pg of pSCll containing the chimeric enu gene or the SFFV enu gene, 1 pg of vaccinia virus DNA, and 10 pg of MuLV LeuVelLeuThrGlnGlnTyrHisGlnLeuLysProLeuGluTyrGluPro salmon sperm DNA/ml of HEPES-buffered saline (Grahamand Van T der Eb, 1973). Stocks of TK- vaccinia virus were prepared in TK-143 cells (Smith and Moss, 1983). To select for recombinants, TK-143 FIG. 1. Comparison of the amino acid sequences of F-MuLV cells were infected with 50-100 plaque-forming units of TK- vaccinia plSE and the corresponding region of SFFV gp52 (from Wolff virus, in the presence of 5-bromo-2-deoxyuridine (25 pg/ml). At 24 h et al., 1983). Only the coding region contained within the AsuIIafter infection, the monolayers were overlayed with 1%low melting KpnI restriction fragment is shown. The first amino acid in both agarose containing 300 pg/ml 5-bromo-4-chloro-3-indolyl-~-~-galacsequences corresponds tothe AsuII restriction enzyme site. The topyranoside. At 4-6 h blue plaques were picked and further purified nonpolar region is underlined. The arrow indicates the cleavage site by 2 additional rounds of plaque purification. Recombinant vaccinia in p15E that generates p12E and the R peptide.
Membrane Stability of a Chimeric gp70/p15E Molecule
to the membrane spanning region and a single base insertion h.p.i. were labeled for 1 h with ['HH]leucine. The sampleswere harvestedandimmunoprecipitatedwithantiseraagainst which alters the reading frame and results in six different amino acids and a premature stop codon at the end of the either gp70 or a synthetic R peptide molecule. Fig. 3 shows membrane spanning region. The chimeric enu gene was in- the wild type (lane A ) and the chimeric(lane B ) glycoproteins serted into the vaccinia virus recombination plasmid p S C l l immunoprecipitated with antisera against gp70. The precurat the SmaI site (Fig. 2). TK-143 cells were infected with sor envelope gene product for the chimericmolecule is about 3300 daltons smaller than the precursor for the wild type vaccinia virus and transfected with pSCl1 containing the molecule. The p15E peptide is present in immune precipitates chimeric enu gene, and vaccinia virusrecombinants were (lane D)when the wild type gp70/p15E sample is reacted with isolated as described under "Materials andMethods." (lane C) in the corresponding Expression of the Chimericgp70/p15E Protein-To identify anti-Rsera,butisabsent immune precipitate with the chimericgp70/p15E. These rethe product of the chimeric gene, TK-143 cells were infected with either a vaccinia recombinant containing the wild type sults show that the chimeric proteinmolecule is lacking the F-MuLV enu gene (Stephens et al., 1986), or with a vaccinia carboxyl-terminal end of the wild type envelope gene which recombinant containing the chimeric enu gene. The chimeric codes for the cytoplasmic tail and theR peptide residues. Intracytoplasmic and surface expression of the chimeric molecule lacks thecytoplasmictail residues foundinthe envelopegene were examined to determine if the lack of normal F-MuLVglycoprotein. As a result, the chimeric PrEnv cytoplasmic tail residues alters the transportof the chimeric molecule should be about 3300 daltons smaller and should gp70/p15E molecule. The defective transport of SFFV gp52 lack thedeterminants recognized by an antiserum to the is thought to be due to the loss of sequences that act as carboxyl-terminal R peptide of MuLV. Cells were infected recognition signalsfor glycoprotein sorting(Srinivasand with the wild type gp70/p15E vaccinia virus recombinant or Compans, 1983b). Fig. 4 shows a similar pattern of intracythe chimericgp70/p15E vaccinia virus recombinant andat 6 toplasmic fluorescence with both the chimeric (panel C) and wild type (panel E ) gp70/p15E glycoproteins. The chimeric gp70/p15E molecule was also found to be expressed at the cell surface (panel D).In contrast, the SFFV gp52 glycoprotein, whenexpressed from a vaccinia recombinant, was found H A to be defective in transport to thecell surface (Fig. 4, G and H). The same percentage of cells was found to express the 1 chimeric molecule a t theirsurfaceas with the wild type Asu 2 ASU 2 Kpn I Kpn I molecule, when profiles of infected cells were compared using afluorescence-activated cell sorter (data not shown). The K-A K m A results indicate that the transport of the chimeric gp70/p15E AmmmDK A Kmolecule is not affected by the lack of a cytoplasmic domain. Polarized Expression of the Chimeric Glycoprotein-Polarized epithelial cells, such as Madin-Darbycanine kidney (MDCK) cells, exhibit apical and basolateral membrane domainsthatareseparated by well-defined tightjunctions. Enveloped viruses were shown to bud exclusively from apical or basolateral membrane domains of polarized cellsin culture (Rodriguez-Boulan and Sabatini, 1978; Herrler et al., 1981; Roth et al., 1983a). Paramyxoviruses and orthomyxoviruses mature at apical surfaces, while retroviruses and rhabdoviruses havebeenshown to mature at basolateral surfaces.
AB Pr ENV" gp70e\ Lac-2 T~
FIG. 2. Constructionof the chimeric MuLV env gene and its insertion into pSCII. The small AsuII-KpnI fragment from P4 (the SFFV genome in pBR322) was ligated to the large AsuII-KpnI 4 --p15E fragment of pRK2 (the F-MuLV enu gene in pCEM3). The recombinant plasmid pMLSF contains the 5' enu sequences from MuLV to FIG. 3. SDS-polyacrylamide gel electrophoresis analysis of the AsuII site from MuLV and the 3' enu sequences from the AsuII of SFFVinsertedintothe vectorpGEM3. The wild type gp70/p15E and chimeric gp70/p15E labeled with to the KpnI site [3H]leucine for 1 h at 6 h.p.i.. A, wild type gp70/p15E and R, chimeric enu gene was inserted into the SmaI site of the vaccinia virus expression plasmid pSCl1. Solid regions indicate plasmid vector chimeric gp70/p15E immunoprecipitated with anti-sera to gp70. C, sequences; vertical lines representSFFV sequences;diagonallines chimeric gp70/p15E and D, wild type gp70/p15E immunoprecipitated R peptide of MuLV. PrEnv is the precursor represent MuLV sequences; open areas indicate the thymidine kinase with antisera to the (TK) gene of vaccinia virus; A, AsuII; Asp, Asp""; H, HindIII; K , envelope protein. The arrows indicatethe position of the p15E KpnI; S, SmaI; X, XbaI; Lac-z, @-galactosidase gene; the asterisk protein. Note that the p15E in lane B is smaller due to the lack of indicates that Asp718 is an isoschizomer of KpnI. the cytoplasmic tail residues.
Membrane Stability of a Chimeric gp70/p15E Molecule
nant (Fig. 5A). However, if the cell monolayers were treated with 30 mM EGTA, a Ca2+ chelating agent thatwill disrupt tight junctions,a characteristic patternof fluorescence typical of a basolaterally expressed glycoprotein was observed at the margins of cells expressing the chimeric or wild type glycoproteins (Fig. 5, C and E). These results indicate that the chimeric glycoprotein molecule is expressed on thebasolateral surface of polarized MDCK cells. Therefore the cytoplasmic domain is notrequired for directional transport of the glycoprotein to basolateral surfaces. Cleavage and MembraneStability of the Chimeric gp70/ pl5E"To determine if the rate of cleavage of the chimeric PrEnv molecule was similar to that of wt gp70/p15E, pulsechase experiments were carried out using CV-1 cells. At 6 h.p.i. CV-1 cells were pulsed for 15 min with 100 pCi of ["SI methionine and chased with medium containing excess cold methionine for 20, 40, 60, 80, and 100 min. As shown in Fig. 6A, gp70 from the chimeric PrEnv molecule can be detected by the end of the 40-min chase. The rates of cleavage of the wild type PrEnv and the chimeric PrEnv were found to be very similar. These results, along with immunofluorescence
FIG. 4. Indirect immunofluorescence of TK-143 cells infected with wild type vaccinia virus( A a n d B ) , chimeric env gene vaccinia virus recombinant (C and D ) , wild type enu gene vaccinia virus recombinant ( E a n d F), or a vaccinia recombinant expressing SFFV gp52 (G a n d H).Fixed monolayers of cells were stained for gp70 with goatanti-gp70antibody, followed by fluorescein-conjugated rabbit anti-goat IgG, and examined for immunofluorescence ( A , C, E, and G ) . Surface immunofluorescence of unfixed monolayers is shown in panels B, D, F, and H.
MDCK cells infected with influenza virus specifically transport hemagglutinin and neuraminidase to the apical plasma membrane (Roth et al., 1983b; Rodriguez-Boulan et al., 1984; Jones et al., 1985), whereas vesicular stomatitis virus G proteinistransportedtothebasolateraldomain (RodriguezBoulan and Pendergast, 1980; Stephens et al., 1986). It was previously shown, usingvaccinia virus recombinants expressing the wt gp70/p15E of F-MuLV, that this glycoprotein is exclusively transported to the basolateral surfacesof MDCK cells (Stephens et al., 1986). We used indirect immunofluorescence to determineif the chimericgp70/p15E molecule is also transported to the basolateral cell surface of MDCK cells. No surface fluorescence was observed on intact MDCK monolayers at6 h.p.i. with the chimericgp70/p15E vaccinia recombi-
FIG. 5. Indirectimmunofluorescence of MDCK cells infected with either the chimeric enu gene vaccinia recombinant or t h e w i l d t y penu e gene vaccinia recombinant in the absence ( A a n d B ) or the presence of EGTA (C, D , E, a n d F).A, absence of fluorescence in intact monolayers of MDCK cells infected withthe chimeric enu gene vaccinia recombinant; E , phase contrast of field shown in A; C,basolateral fluorescence on MDCK cells infected with the chimeric enu gene vaccinia recombinant and treated a t 6 h.p.i. with 30 mM EGTA for 30 min, followed by staining for gp70;D,phase contrast of field shown in C; E , basolateral fluorescence on MDCK cells infected with the wild type enu gene vaccinia recombinant and treated a t 6 h.p.i. with 30 mM EGTA for 30 min, followed by staining for gp70; F, phase contrast of field shown in E.
Chimeric of a
gp70/p15E Molecule shows that the majority of w t gp70 is cell associated at the 20-h time point, with small amounts present in the supernatant. Unlike wt gp70, very little of the gp70 of the chimeric glycoproteins remained cell associated a t 20 h (Fig. 7B) with the majority being released into the medium. These results indicate that the lack of a cytoplasmic domain leads to instability of the glycoprotein on the cell surface.
FIG. 6. Pulse-chase analysisof immunoprecipitated proteins that were either cell associated (A) or released into the media ( E ) .TK-143 cells were infected with either the chimeric enu gene vaccinia recombinant ( C ) or the wild type enu gene vaccinia recombinant ( W) and pulsed for 15 min at 6 h.p.i. with [:''S]methio-
We have constructed a recombinant vaccinia virus expressing a chimeric envelope gene derived from the 5' end of the F-MuLV envelope gene and the3' end of the SFFV envelope gene. Sequence analysis of a variety of SFFV clones shows nine. The labeled samples were harvested after chase periods of 0,20, that the threemajor mutations which differentiate the SFFV 40,60, 80, and 100 min. A, cell-associated proteinswere immunoprecipitated with anti-sera against gp70. R, the media from the samples envelope gene from the MuLV and mink cell focus-forming in A wascollected and immunoprecipitated with anti-sera against virus and envelope genes are a large deletion, and two insergp70 a t the specified time points. The PreEnv of the chimeric enu tions of six and one bases(Wolff et al., 1983; Clark and Mak, gene protein is about 3300 daltons smaller due to the lack of cyto- 1984). We have analyzedthe effects of two of these mutations, plasmic tail residues. Gp70 is seen at 80 and 100 min for the chimeric i.e. insertions of six and one bases, by incorporating them sample in p a n e l H. The asterisk indicates a protein which is nonspe- into theenvelope gene of F-MuLV. These changes in the enu cilically precipitated. gene of MuLV resulted in the addition of 2 leucine residues to the membrane spanningregion and in the lack of residues CELLS MEDIA CELLS MEDIA which normally form the cytoplasmic tail. 1 2 3 14 2 3 41 . 2 3 14 2 3 4 . The intracellular transportof membrane glycoproteins and A B .-. their subsequentlocalization in differentorganelles is thought to be mediated by transport vesicles (Rothman and Fine, 1980) and specific recognition signals are thought to be involved in the packaging into transport vesicles of glycoproteins destined for subsequent localization at particular compartments of the cell (Blobel, 1980; Pearse and Bretscher, 1981). The defective transport of SFFV gp52 may be due to the lack of sorting signals. We analyzed the transport of the FIG. 7. SDS-polyacrylamide gel electrophoresis analysis of cells infected with the wild type env gene vaccinia recombi- chimeric gp70/p15E in CV-1 cells and in a polarized MDCK nant ( A ) or the chimeric enu gene vaccinia recombinant ( E ) . cell line, to determine whether sorting signals are located in Cells wereinfected and at 6 h.p.i. werelabeledfor 1 h with ["HI the cytoplasmic tail region of F-MuLV gp70/p15E. The chiglucosamine. The samples were collected at: 1 , O h 2, 1 h; 3, 3 h; and meric molecule, which lacks the cytoplasmic domain, was 4, 20 h following the labeling period. The envelope proteins present transported to the surfacesof CV-1 cells, and the basolateral in the cellular fraction and the supernatant medium were analyzed surface of MDCK cells, indicating that the lack of a cytoby immunoprecipitation with anti-sera against -70. The majorityof the wild type gp70 is seen to remain cell associated at 20 h, while the plasmic tail did not block thetransport of thechimeric molecule. Theseresultsareconsistentwith several other majority of the chimeric gp70 is released by 20 h. reports that the cytoplasmic domains of othermembrane studies,indicatethatthechimeric gp70/p15E molecule is glycoproteins carry no signals which alter the rate or effiefficiently cleaved and transported to thecell surface. Witte ciency of the glycoprotein transport. Gething et al. (1985) and Wirth (1979) showed that the proteolyticcleavage of the reported that there is no dominant signal within the cytoplasmic domain of the HA glycoprotein of influenza which MuLV PrEnv molecule is a late step in maturation and is determines the rate of transport for the glycoprotein. Altering closely timedtothecarbohydrate modification. It isnot known whether this cleavage occurs as themolecule is trans- the cytoplasmic tail of influenza neuraminidase, an apical ported from the rough endoplasmic reticulum to thecis-Golgi protein, did not disrupt its sorting (Jones et al., 1985). McQueen et al. (1986) reported thata chimeric protein containing compartment or shortly after entering the cis-Golgi. the external domain of the influenza HA molecule and the T o determinethe effect of the cytoplasmic domainon stability of the protein on the cell surface, the supernatants transmembrane and cytoplasmic domains of the vesicular from this experiment were also collected and examined for stomatitis virusG protein (a basolaterally expressed protein) the presence of gp70. The gp70 cleavage product from the is transported to and localized on the apical membrane of chimeric PrEnv molecule was detected in the culturefluid at polarized MDCK cells. Similarly, McQueen et al. (1987) demthe 80-min chase period (Fig. 623). No wt gp70 was seen in onstrated that a chimera containing thecleavable leader and the supernatant fractions; however, trace amounts could be ectodomain of vesicular stomatitis virus G fused to the cytodetected in other pulse-chase experiments (data not shown). plasmic and transmembrane domainsof HA is transported to Pinter and Honnen(1985) reported that thecell surface form the basolateral surface of polarized cells. These results indifor apicaltransport of HA and for basolateral of the Friend SFFV enu protein, gp65, is efficiently released cate that signals from cells after a 20-h chase. Since the chimeric gp70 could transport of G reside in the respective ectodomains and not the transmembrane or cytoplasmic domains. Similar results be detected in culture fluids by 80 min, we examined later time points to determine whether the majority of the chimeric have been obtained using progressive deletionmutants within cytoplasmic domains of the influenza gp70 was released into the medium. CV-1 cells were infected the transmembrane and with the vaccinia recombinant and pulsed for 1 h at 6 h.p.i. HA (Doyle et al., 1986). Roman and Garoff (1986) also found p62/ with 200 pCi of [RH]glucosamine. Cell-associated and super- that thecytoplasmic domain of the Semliki Forest virus natant fractions were collected after 1, 3, and 20 h. Fig. 7A E2 protein does not contain the information essential for
of a Chimeric gp701p15E Molecule
directed transport to the basolateral membrane of MDCK cells. Similarly, the removal of the cytoplasmic tail from the murine histocompatibility antigendid not prevent the expression of the protein on the surface of cells (Zuniga and Hood, 1986). Furthermore, when the coding sequence for the cytoplasmic tail of F-MuLV are inserted in place of the 3’ end of the SFFV enu gene, the resulting protein is still defective in transport to the cell surface (Srinivas et al., 1987). Taken together, these results suggest that the transport defect for SFFV gp52 is due to the large 585-base deletion within the enu gene, and not to the lack of cytoplasmic tail. The large deletion may have resultedin the loss of sorting signals. However, it is also possible that the large deletion resulted in a molecule which does not fold correctly, or which does not assemble into an oligomeric structure (Gething et al., 1986), which could alter the rate of transport ofgp52 from the endoplasmic reticulum to the Golgi complex. The major effect of the lack of a cytoplasmic domain may be the loss of stability in association of the glycoprotein with membranes. The carboxyl-terminal sequences of the F-MuLV p15E contain a domain of 30 uncharged residues followed by an additional 32 residues which include a number of charged amino acids (Shinnick et al., 1981). Pinter and Honnen (1985) proposed that deletion of the 32 residues results in areduction in efficiency of the stop transferfunction of this domain and therefore accounts for the spontaneous release of gp65 from the plasma membrane. They reported that gp65 was released from SFFV-infected nonproducer rat cell lines, although the amount of gp65 released into the medium varied, depending upon the different cloned cell lines analyzed. We found that the vaccinia virus recombinant containing the chimeric envelope gene released the chimeric gp70 molecule intothe medium as early as 80 min following the radioactive pulse, and large amounts of chimeric gp70 were found in the supernatant by 20 h, with very little remaining cell associated. The release of the chimeric glycoprotein from infected cells may be due to thelack of charged residues on the cytoplasmic side of the membrane spanning domain, indicating that these residues may confer stability on the association of the wild type proteinwith the plasma membrane.It hasbeen proposed that the release of gp65 may be involved in the induction of erythroleukemia (Pinter and Honnen, 1985). To further understand therole of the envelope glycoprotein in the induction of disease, we intend touse the chimeric construction to study its effect on the disease process by infecting adult mice with F-MuLV containing the chimeric enu gene. Acknowledgments-We thank Dr. D. Linemeyer and Dr. A. Oliff for kindly supplying the plasmids P4 and p231, and Dr. B. Moss for providing the vaccinia virus expression vector, pSC11, and therecombinant vaccinia virus expressing F-MuLV envelope proteins. The rabbit antiserum prepared against a synthetic peptide with R antigenic determinants was kindly provided by Dr. R. Lerner. REFERENCES Amanuma, H., Katori, A., Obata, M., Sagata, N., and Ikawa, Y.(1983) Proc. Natl. Acad. Sci. U. S. A. 8 0 , 3913-3917
Blobel, G. (1980) Proc. Natl. Acad. Sei. U. S.A. 77, 1496-1500 Clark, S. P., and Mak, T. W. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,5037-5041 Clark, S. P., and Mak, T. W. (1984) J. Virol. 50, 759-765 Dickson, C., Eiseman, R., Fan, H., Hunter, E., and Teich, N. (1984) in RNA Tumor Viruses (Weiss, R., Teich, N., Varmus, H., and Coffin, J., eds) pp. 513-648, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY Doyle, C., Sambrook, J., and Gething, M. J. (1986) J. Cell Biol. 103, 1193-1204 Gething, M. J., Doyle, C., Roth, M. G., and Sambrook, J. (1985) Curr. Top. Membr. Tramp. 2 3 , 17-41 Gething, M. J., McCammon, K., and Sambrook, J. (1986) Cell 46, 939-950 Graham, F. L., and Van der Eb, A. J. (1973) Virology 52,456-467 Herrler, G., Nagele,A., Meier-Ewert, H., Bhown, A. S., and Compans, R.W. (1981) Virology 1 1 3 , 439-451 Jones, L. V., Compans, R. W., Davis, A. R., Bos, T. J., and Nayak, D. P. (1985) Mol. Cell Bwl. 5 , 2181-2189 Laemmli, U.K. (1970) Nature 227,680-685 Linemeyer, D. L., Menke, J. G., Ruscetti, S. K., Evans, L. H., and Scolnick, E. M. (1982) J. Virol. 4 3 , 223-233 Machida, C. A., Bestwick, R. K., and Kabat, D. (1984) J. Virol. 49, 394-402 McQueen, N., Nayak, D. P., Stephens, E. B., and Compans, R. W. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,9318-9322 McQueen, N. L., Nayak, D. P., Stephens, E. B., and Compans, R. W. (1987) J. Biol. Chem. 2 6 2 , in press Pearse, B. M. F., and Bretscher, M. S. (1981) Annu. Rev. Biochem. 50,85-101 Pinter, A., and Honnen, W. J. (1985) Virology 143, 646-650 Rigby, P. W. J., Dieckman, M., Rhodes, C., and Berg, P. (1977) J. Mol. Biol. 137, 237-251 Rodriguez-Boulan, E. J., and Sabatini, D. D. (1978) Proc. Natl. Acad. Sci. U. S. A. 76,5071-5075 Rodriguez-Boulan, E. J., and Pendergast, M. (1980) Cell 20,45-54 Rodriguez-Boulan, E. J., Paskiet, K. T., Salas, P. J. T., and Bard, E. (1984) J. Cell Biol. 9 8 , 308-319 Roman, L. M., and Garoff, H. (1986) J. Cell Biol. 103, 2607-2618 Roth, M. G., Srinivas, R. V., and Compans, R. W. (1983a) J. Virol. 45, 1065-1073 Roth, M. R., Compans, R. W., Giusti, L., Davis, A. R., Nayak, D. P., Gething, M.-J., and Sambrook, J. (1983b) Cell 33,435-443 Rothman, J. E., and Fine, R. E. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,780-784 Ruta, M., and Kabat, D. (1980) J. Virol. 35,844-853 Ruta, M., Bestwick, R., Machida, C., and Kabat, D. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,4704-4708 Shinnick, T. M., Lerner, R.A., and Sutcliffe, J. G. (1981) Nature 293,543-548 Smith, G. L., and Moss, B. (1983) Genet (Amst.) 2 6 , 21-28 Srinivas, R. V., and Compans, R.W. (1983a) J. Biol.Chem. 2 5 8 , 14718-14724 Srinivas, R. V., and Compans, R. W. (1983b) Virology 125,274-286 Srinivas, R.V., Kilpatrick, D. R., and Compans, R. W. (1987) J . Virol., in press Stephens, E. B., Compans, R.W., Earl, P., and Moss, B. (1986) EMBO J. 5 , 237-245 Sutcliffe, J. G., Shinnick, T. M., Green, N., Liu, F. T., Niman, H. L., and Lerner, R. A. (1980) Nature 287,801-805 Witte, 0. N., and Wirth, D. F. (1979) J. Virol. 2 9 , 735-743 Wolff, L., Scolnick, E., and Ruscetti, S. (1983) Proc. Natl. Acad. Sci. U. S. A. 8 0 , 4718-4722 Zuniga, M. C., and Hood, L. C. (1986) J. Cell Biol. 102, 1-10