Genetic and Siaiylation Sources of Heterogeneity of the Murine ...

Vol. 254, No. 4, Issue of February 25, pp. 1340-1348, 1979 Printed m U.S.A.

Genetic and Siaiylation Sources of Heterogeneity of the Murine Leukemia Virus Membrane Envelope Glycoproteins gp69/71*

(Received for publication, January 24, 1978)

Mark J. Murray and David Kabat

From the Department of Biochemistry, School of Medicine, University of Oregon Health Sciences Center, Portland, Oregon 97201

We have analyzed the causes for heterogeneity of the membrane envelope glycoprotein of murine leukemia viruses (MuLVs). When analyzed by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, this glycoprotein frequently separates into two components which have been termed gp69/71 because they migrate in the gels with apparent molecular weights of 69,000 and 71,000, respectively. Cloning experiments with MuLV produced by cultured Eveline cells indicates that gp69 and gp71 are encoded by separate closely related viral genomes which may frequently become incorporated into heterozygous virus particles. This and other evidence demonstrate that the size difference between gp69 and gp71 is caused by a difference in the amino acid sequences of their polypeptide chains. The gp69 and gp71 glycoproteins also have extensive isoelectric point microheterogeneity which can be reduced by removal of sialic acids with neuraminidase. Analyses of n-[3H]glucosamine-labeled glycopeptides by anion exchange chromatography, by gel filtration, and by high voltage paper electrophoresis indicate that the envelope glycoproteins are glycosylated at several sites. It appears that the glycoproteins contain several complex oligosaccharides which are heterogeneously sialylated and at least two other relatively small neutral oligosaccharides. It is concluded that heterogeneity of the envelope glycoprotein of MuLVs can be caused by at least two factors, a genetic heterogeneity caused by multiplicity of viral genomes and a variability in the numbers of sialic acid residues attached to oligosaccharides.

The major protein component on the surface membrane of mammalian RNA tumor viruses is a glycoprotein (l-4) which plays a major role in virus infections and possibly also in malignant transformation. Interaction of the glycoprotein with a cellular receptor is required for penetration of the virus into cells (5), and this interaction plays a role in determining many of the host range and interference properties of the murine leukemia viruses (5-7). Furthermore, infected cells contain this glycoprotein on their plasma membrane (l-4,8-10) where it is believed to be involved in virus budding from the cell surface (11, 12). In addition, normal uninfected mice contain a family of genes for glycoproteins closely related to the MuLV’ envelope glycoproteins (13-17). These endogenous

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

’ The abbreviations used are: MuLV, murine leukemia virus; F- MuLV, Friend murine leukemia virus; EIIV, MuLV released from Eveline cells; B4Sc-1, the B4 clone of EIIV growing in SC-I cells;

glycoproteins may be produced in different tissues at specific stages of differentiation (13,18-20); and uninfected mice often produce antibodies against these glycoproteins (21-24). Fi- nally, recent evidence suggests an important role for the MuLV envelope glycoprotein in leukemogenesis (25-27).

Despite their biological importance, the molecular properties of these glycoproteins are poorly understood. The membrane envelope glycoprotein of MuLVs and of other mammalian C-type retroviruses frequently separates into two polydisperse components when analyzed by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (28-30). These components have been termed gp69/71 because they migrate in the gels with apparent molecular weights of 69,000 and 71,000, respectively (28-31). Several studies have indicated that these two components could not be distinguished serologically or by tryptic peptide analysis (13,32). Although their glycopeptides have not been analyzed, it has been proposed that the size heterogeneity of gp69/71 may be caused by a difference in amount of carbohydrate added onto a single polypeptide chain (28, 33).

In this paper we describe an analysis of MuLV gp69/71 produced by the Eveline cell line. This cell line was derived by infecting STU mouse cells with Friend MuLV (34). These cells produce a large amount of MuLV (35) and have been used extensively for purification (36) and immunological studies (37-39) of the MuLV envelope glycoprotein. We demonstrate that MuLV prepared from these cells has the gp69/71 doublet, whereas virus cloned from these cells produces only one glycoprotein component. We suggest that the size heterogeneity of gp69/71 may generally be caused by a multiplicity of viral genomes within many of the established stocks of MuLV. These different viral genomes may arise by deletion of a portion of the envelope glycoprotein gene. Furthermore, several lines of evidence suggest that the glycoproteins from the cloned and uncloned viruses have isoelectric point microheterogeneity caused partly by heterogeneous sialylation of carbohydrate side chains.

EXPERIMENTAL PROCEDURES

Cells and Virus

A subline of Eveline cells (Eveline II cells) were kindly provided by D. Bolognesi, Duke University Medical Center, Durham, N. C. These cells, which were originally derived by infection of STU mouse cells with Friend virus (34), produce large amounts of MuLV and negligible amounts of spleen focus forming virus (35, 40). The cells were grown as suspension cultures in Dulbecco’s modified Eagle’s medium

gp69/71, MuLV envelope glycoproteins with apparent molecular weights of 69,000 and 71,000; gPr89/91”“, precursors of MuLV envelope glycoproteins (designated gp89/91 in some figures); Pr72/73”“‘, unglycosylated precursors of MuLV envelope glycoproteins; NEPHGE, nonequilibrium pH gradient electrophoresis method of O’Farrell et al. (47); NRK, normal rat kidney.

1340

by guest on February 16, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Murine Leukemia Virus Glycoproteins 1341

(GIBCO) containing 10%~ complement inactivated (30 min at 56°C) fetal calf serum (GIBCO). Sixty-milliliter cultures growing in 250.ml Erlenmeyer flasks (Corning) were passaged at 2- to 3.day intervals. The cell concentration was maintained between 6 x 16’ and 3 x 10” cells/ml.

The NRK and SC-~ cells were provided by Contract E-73-2001- NO1 within the Special Virus-Cancer Program, National Institutes of Health, United States Public Health Service through the courtesy of Jack Weaver at the Cell Culture Laboratory, University of California School of Public Health. The SC-I cell line was derived from a feral mouse embryo (41). This line shows no Fv-I restriction to N- or B- tropic murine leukemia viruses and is apparently free of endogenous virus expression. The cells were maintained in monolayer culture with McCoy’s modified 5A medium (GIBCO) supplemented with 10% complement inactivated fetal calf serum and antibiotics (GIBCO). The NRK cells were maintained in monolayer culture with Eagle’s minimal essential medium (GIBCO) supplemented with 10% comple-

virions labeled with n-[“HJglucosamine showed that all detectable label (>95%) was incorporated into the envelope glycoproteins gp69/71. In agreement with previous studies, the Eveline cells were relatively active in MuLV synthesis (35) and incorporation of D- [“Hlglucosamine into glycoproteins (37-39). In contrast, the B4Sc-I cells incorporate only small amounts of n-[“Hlglucosamine into glycoproteins.

Cellular Labeling and Immune Precipitations

Procedures for labeling cells with L-[““S]methionine and for secondary immune precipitation of cell extracts with monospecific anti- sera to gp69/71 have been described previously (40). Antiserum to Eveline virus gp69/71 was generously donated-by J. Collins, Duke University, Durham, N. C., and was shown to be monosnecific as described previously.

Cell Surface Radioiodination ment inactivated fetal calf serum and antibiotics.

The Eveline virus was cloned on SC-1 cells, Medium from a loga- Cell surface labeling with [‘YI]iodine was performed at 0°C by a

rithmically growing Eveline cell culture was centrifuged at 4,000 rpm modification (10) of Vitetta et al. (45). Growing cells were used for

to remove the cells and at 10.000 ram to remove debris and was then labeling.

diluted (125 times) with fresh culture medium. The dilution reduced viral infectivity from 250,000 foci/ml to 2,000 foci/ml. Viral infectivity was determined using the S’L- assay (42). The medium was removed from an SC-~ culture seeded the previous day with 2 x IO’ cells (25 cmL T-flask, Falcon Plastics). The cell monolayer was incubated with 1 ml of DEAE-dextran (25 kg/ml in culture medium, Sigma) at 37°C for 30 min. The DEAE-dextran was removed and 0.5 ml of the dilute Eveline virus inoculum was put onto the cells and incubated for 90 min at 37’C. Following the incubation period, the inoculum was removed and the cells were dispersed with trypsin/EDTA solution (GIBCO) and diluted with medium to 200 cells/ml. Then 40 cells (0.2 ml) were seeded into each well of a microtiter plate (Falcon Plastics) and incubated at 37°C in a 5% CO, atmosphere. When the cells in the microtiter wells were 60 to 70% confluent, the medium was removed for S’L- assay of viral infectivity and fresh medium was added to the cells. Cells were maintained in”the wells an additional 5 to 6 days, until the results of the S’L- assay were known. Then cells from virus- producing wells were grown up to 25-cmL cultures and were maintained like the uninfected SC-1 cells. These procedures were designed to give multiplicities of infection of 5 x 10mJ, although in practice they appeared to be somewhat lower (5 x 10m4).

The EIIVSc-1 line was constructed simply by infecting SC-I cells with whole (uncloned) Eveline virus essentially as described above. Medium from logarithmically growing Eveline cells was clarified bv centrifugation, adjusted to 8 pg/ml of polybrene (Aldrich), and inoc- ulated onto an SC-1 monolayer. None of the steps designed to yield a clone were used, i.e. dilution of culture medium or dispersal of infected cells.

The B4NRK cell line was constructed by infecting NRK cells with the virus produced by the B4Sc-1 line. B4Sc-1 is an Eveline virus clone growing in SC-1 cells.

The Friend virus-induced erythroleukemia cells (line F4-6/K) (43, 44) were generously donated by W. Ostertag, Max-Planck Institut fur Experimentelle Medizin, Gottingen, Germany. Friend virus (original strain) was obtained from the American Type Culture Collection.

Labeling of Virion Proteins and Purification of Virus

Viral proteins were labeled with L-[““Slmethionine, L-[“Hlleucine or D-[3H]glucosamine (New England Nuclear). Amino acid labeling was performed by incubating virus-producing cells in minimum essential medium (GIBCO) containing 10% dialyzed fetal calf serum but lacking the appropriate amino acid and supplemented with the iso- topically labeled amino acid. The labeled amino acids were included at concentrations of 15 to 20 @/ml. Before labeling, the cells were washed with phosphate-buffered saline (GIBCO). The cells were then incubated with the radioactive medium at 37°C in 5% CO2 for 2 h, and the radioactive medium was then removed and replaced with fresh complete medium. The cultures were further incubated for 18 to 24 h, and the virus was harvested. D-[“H]Glucosamine labeling was accomplished simply by adding it to complete medium (20 @.X/ml) and incubating the cells for 8 to 18 h, after which time the virus was harvested.

The procedures for harvesting and purification of virus have been described previously (40). Simply stated, the virus was pelleted from the culture medium by centrifugation and was then centrifuged to equilibrium in a 15 to 60% sucrose density gradient. Analysis of the

Polyacrylamide Gel Electrophoresis

The procedures for one-dimensional sodium dodecyl sulfate/8 M urea-polyacrylamide gel electrophoresis have been described previously (40). Two-dimensional polyacrylamide gel electrophoresis was performed essentially as described by O’Farrell (46) with minor modification in the first dimension: 9 M urea (ultrapure, Schwarz/Mann), 1 M electrode solutions, and 2% ampholytes (LKB) (pH 3.5 to lO/pH 9 to Il/pH 4 to 6/pH 5 to 7 in the ratio 14:2:1:1). Two-dimensional nonequilibrium pH gradient electrophoresis was performed using pH range 3.5 to 10 Ampholines as described by O’Farrell et al. (47). The NEPHGE system was designed to maximize resolution of basic proteins. After electrophoresis, the gel slabs were fixed overnight in 12.5% trichloroacetic acid and were processed for fluorography (48).

Neuraminidase Treatment

The protease-free neuraminidase (mucopolysaccharide N-acetyl- neuraminylhydrolase (EC 3.2.1.18) from Vibrio cholerae) was obtained from Behringwerke (Marburg/Lahn, Germany). D-[,‘H]Gluco- samine-labeled virus (20,000 to 40,000 cpm) was suspended in 10 mM Tris-HCl, 100 mM NaCl, 1 mM EDTA, pH 7.4, and was incubated at 37“C for 2 h in a shaking water bath with at least 10 units of enzvme (500 units/ml). Equal amounts of neuraminidase were added tdthe samples at the beginning of the incubation and again after 1 h of incubation. The samples were then rapidly frozen. Control samples were incubated in parallel without neuraminidase. Selected incubations included additional Ca’+ ion. The per cent of D-[ ‘Hlglucosamine label in the radioactive glycoprotein which was incorporated after conversion to sialic acids was determined by quantitating the amount of radioactivity which became soluble in 5% trichloroacetic acid after neuraminidase treatment.

Glycopeptide Analyses

DEAE-Sephadex A-25 Chromatography-After the neuraminidase treatment described above, the D-[“Hlglucosamine-labeled samples were lyophilized and the residue was solubilized in 200 el of 0.05 M ammonium bicarbonate, pH 8.5. Tosyl phenylalanyl chloromethyl ketone-treated trypsin (25 pg) (Worthington) was added and the samples were incubated at 37°C for 24 h. The resulting digest was lyophilized and the residue was dissolved in 2 ml of 50 mM Tris-HCl, pH 8.5. The tryptic glycopeptide samples were applied to a DEAE- Sephadex A-25 (Pharmacia) column (0.9 x 20 cm) and were eluted with a linear 600 ml 0 to 0.3 M NaCl gradient (49). The flow rate was approximately 1 ml/min and 1.8ml fractions were collected. Each fraction was added to 15 ml of Aquasol (New England Nuclear) and was counted in a Packard Tri-Carb liquid scintillation spectrometer. The peak which emerged from this column at Fraction 20 was identified as free sialic acid by cochromatography with N-[l’C]acetylneuraminic acid (Amersham/Searle).

Sephadex G-50 Chromatograph-The fractions constituting the runoff peak from the DEAE-Sephadex columns were pooled and lyophilized. The lyophilized residues were dissolved in 100 11 of 0.05 M ammonium bicarbonate, 5% isopropyl alcohol (v/v). These samples were pronase (Calbiochem) treated by addition of 200 ~1 of 0.1 M Tris- HCl, 1 mM CaC12, pH 7.8, containing 1 mg/ml of pronase, overlain with toluene, and incubated at 37°C for 72 h. At 24 and 48 h, an


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

1342 Murine Leukemia

additional 200 ~1 of the pronase solution was added. At 72 h, the incubations were stopped and the samples were lyophilized. The lyophilized samples were solubilized in 0.5 ml of 0.05 M ammonium bicarbonate, 5% isopropyl alcohol. They were then applied to a Sephadex G-50 superfine (Pharmacia) column (0.9 x 95 cm) and eluted with the same buffer at a flow rate of 2.5 ml/h. Each fraction (0.35 ml) was added to 10 ml of Aquasol and was counted in a Packard Tri-Carb liquid scintillation spectrometer. The column-excluded volume, determined with blue dextran, was 24 ml and the included volume, determined with phenol red, was 70 ml.

High Voltage Paper Electrophores&-n-[3H]Glucosamine labeled virus (25,000 cpm) was neuraminidase-treated as described above and lyophilized. The neuraminidase and control samples were then resus- pended in 1 ml of 0.1 M Tris-HCl, 1 mM CaCl2, pH 7.8, containing 1 mg/ml of pronase and overlain with toluene. The samples were incubated at 37°C for 72 h. Additional I-mg amounts of pronase were added to the incubation mixtures at 24 and 48 h and the samples were lyophilized at 72 h. The samples were then redissolved in electrophoresis buffer (pyridine/acetic acid/water, 10:0.4:89.6) and aliquots were electrophoresed at 3,000 V for 3 h on Whatman No. 3MM paper (3 x 60 cm). Following electrophoresis, the paper strips were cut into 0.5~cm sections and incubated in scintillation vials with 0.5 ml of protosol:toluene (1:2, New England Nuclear) at 25°C for 30 min. After addition of 5 ml of toluene/acetic acid (Baker scintillation grade toluene, 0.3% 2,5diphenyloxazole (PPO), 0.03% 1,4-bis[2-(5-phenylox- azoly)]benzene (POPOP), 0.1% glacial acetic acid), each section was counted in a Packard Tri-Garb liquid scintillation spectrometer. The position of the electroosmotic front was determined using the migration position of [3H]water (New England Nuclear) as standard.

RESULTS

Genetic Heterogeneity of Eveline Virus gp69/71-In order to study the synthesis of gp69/71, a suspension culture of Eveline cells was pulse-labeled with L-[35S]methionine and was subsequently chased with a large excess of nonradioactive methionine. Cell samples were harvested at various times during this labeling procedure. Fig. 1 shows an electrophoretic analysis of the radioactive proteins which were precipitated from the cell extracts with monospecific antibody to gp69/71. Consistent with previous studies (30, 33, 40, 50), the two gp69/71 glycoproteins are derived by processing two larger precursors which have apparent molecular weights of 89,000 and 91,000. The gPr89/91”“” precursors contain core carbohydrates but lack the terminal sugars fucose and sialic acid which occur in gp69/71 (30,50). Our results show that gPr89”“” and gPr91’“” occur in an approximately constant ratio throughout the labeling period, suggesting that they are not related to each other in a precursor-product manner. Further- more, they are rapidly formed and their unglycosylated precursor(s) are not detected. Core carbohydrates are often rapidly added in N-glycosidic linkage onto nascent polypeptides as they enter the extracytoplasmic side of the rough endo- plasmic reticulum (51-54), whereas fucose and sialic acid termini are added after glycoproteins have migrated into the Golgi apparatus (51). While gPr89”“” and gPr91”“” are present in equal amounts in this experiment, the immune precipitations sometimes yield unequal amounts of these precursors.

Evidence that the two components of gp69/71 are encoded by different viral genomes was obtained using cloned Eveline virus. Fig. 2 shows an experiment in which one virus clone (B4) was grown in SC-1 fibroblasts and was analyzed as described above. SC-1 cells infected with uncloned Eveline virus (Lanes A to C) synthesize the gPr89/91”“” and gp69/71 doublets as expected, but cells infected with the cloned B4 virus (Lanes D to F) synthesize only the single glycoproteins gPr91”“” and gp71. Normal rat kidney cells infected with the B4 virus (Lanes G to I) also synthesize only the gPr91”“” and gp71 glycoproteins. These results suggest that gPr91”“” and gPr89”“” are precursors of gp71 and gp69, respectively. Fur- thermore, the two virion glycoproteins in Eveline cells are

Virus Glycoproteins

A B CDEFGH I J

-bp 69/71

FIG. 1. Pulse-chase analysis of gp69/71-specific proteins in Eveline cells. Eveline cells (1.4 x 10” cells/ml) were pulse-labeled with 50 @i/ml of L-[““Slmethionine for 30 min in methionine-free minimal essential medium. At the conclusion of the pulse period the cells were chased by the addition of unlabeled methionine for 2 h. Samples taken at various times were immune-precipitated by primary antiserum precipitations and were analyzed in 10% polyacrylamide slab gels containing 0.1% sodium dodecyl sulfate and 8 M urea as described previously (40). Lanes A to J contain immune precipitates obtained after: A, 1-min pulse; B, 5-min pulse; C, lo-min pulse; D, 20-min pulse; E, 30-min pulse; F, 30-min pulse and 15-min chase; G, 30-min pulse and 22-min chase; H, 30-min pulse and 30-min chase; 1, 30-min pulse and 60-min chase; J, 30-min pulse and 120-min chase. The virus- specific proteins are indicated at the right of the gel. The current nomenclature for gp89/91 is gPr89/91’““. t-[?S]Methionine-labeled Eveline virus was electrophoresed in another well (not shown) to determine the positions of migration of gp69/71. The gp89/91 components were visible in Lane A when the film exposure was prolonged.

A 8 C D E F G H I J K

-F!m+ -Es+ w SC-I NRK

FIG. 2. Pulse-chase analysis of gp69/71specific proteins of Eveline virus in SC-~ cells and clone B4 virus in SC-~ and NRK cells. Mono- layer cultures (75 cm”) of virus-infected SC-~ and NRK cells at 50 to 70% confluencey were pulse-labeled for 30 min with 20 $.X/ml of L- [““Slmethionine in 3 ml of methionine-free minimal essential medium. At the conclusion of the pulse period the cells were chased by addition of unlabeled methionine (10 ml of complete minimal essential medium) for 30 min and for 2 h. Samples were collected at the end of the pulse (30 min) and both chase (30 and 120 min) periods. Uninfected cells were pulse-labeled for 60 min. The labeled cells were then lysed, the cell extracts were immune-precipitated by secondary antiserum precipitations using monospecific antibody to gp69/71, and the immunoprecipitates were analyzed by electrophoresis in 10% polyacrylamide slab gels containing 0.1% sodium dodecyl sulfate and 8 M urea as previously described (40). Lanes A to C are immune precipitates from whole Eveline virus-infected SC-~ cells (EIIVSc-1) obtained after: A, 30-min pulse; B, 30-min pulse and 30-min chase; C, 30-min pulse and 120-min chase. Lanes D to Fare immune precipitates from B4 virus clone-infected SC-~ cells (B4Sc-1) obtained after: D, 30-min pulse; E, 30-min pulse and 30-min chase; F, 30-min pulse and 120-min chase. Lanes G to I are immune precipitates from B4 virus clone- infected NRK cells (B4NRK) obtained after: G, 30-min pulse; H, 30- min pulse and 30-min chase; 1,30-min pulse and IfO-min chase. Lane J is an immune precipitate from a 60-min pulse of uninfected SC-~ cells. Lane K is an immune precipitate from a 60-min pulse of uninfected NRK cells. The current nomenclature for gp89 and gp91 is gPr89’“” and gPr91’““.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


apparently produced by different viral genomes which are separable by cloning and encode for the synthesis of distinct polypeptide chains. It is also apparent that the size difference between gPr89’“” and gPr91’“” and thus between gp69 and gp71 cannot be the consequence of a processing heterogeneity (e.g. glycosylation or proteolysis) within the host cell.

An analysis of L-[35S]methionine-labeled virion proteins by two-dimensional, isoelectric focusing/sodium dodecyl sulfate- polyacrylamide gel electrophoresis (46) is shown in Fig. 3. The gp69/71 glycoproteins separate in the isoelectric focusing (hor- izontal) dimension with a broad “string-of-beads” appearance. These proteins were identified as gp69/71 by labeling with n-[3H]glucosamine and also by immune precipitation with antibody to gp69/71. Fig. 3 shows that the virion glycoproteins from Eveline virus (Panel A) separate in the sodium dodecyl sulfate electrophoresis (vertical) dimension as two rows of spots which differ slightly in size, whereas the cloned B4 virus (Panel B) contains only the larger size class of envelope glycoproteins. These results support the above conclusions based on immunoprecipitation of intracellular proteins (Figs. 1 and 2). As described below, the row of gp69 glycoproteins appears to have slightly more basic components than the gp71 glycoproteins. Furthermore, the relatively acidic components often appear to have slightly larger molecular weights than the basic components.

Using these techniques, we have also analyzed three other virus clones derived from the Eveline virus. One of these was like the B4 clone and synthesized only the gPr91en” precursor and the gp71 envelope glycoprotein. The other two clones produced the doublets gPr89/91’“” and gp69/71. Clones such as the latter could have occurred if virus aggregates were present in the preparation of Eveline virus used for cloning. However, an electron microscopic examination of this virus did not show any aggregates. Alternatively, such clones would occur if many of the virus particles released from Eveline cells are heterozygotes which contain one genome coding for gp69 and a second coding for gp71. These possibilities will be discussed below.

Additional Evidence for Structural Difference between the Polypeptide Chains of the Envelope Glycoproteins-The results described above indicate that distinct genes specify gp69 and gp71 and that the encoded polypeptide chains must differ in their amino acid sequences and possibly also in their sizes. Additional evidence for a structural difference between the

-9P71 -gp 69

-gP7l

i

FIG. 3. Two-dimensional polyacrylamide gel electrophoresis of Eveline virus gp69/71 and clone B4 virus gp71. Virion proteins were labeled with L-[3H]leucine as described under “Experimental Proce- dures.” Purified virus was denatured and electrophoresed in two dimensions essentially as described by O’Farrell (46). Only the gp69/71 region of the resulting autoradiograms is shown. Panel A is Eveline virus gp69/71 and Panel B is clone B4 virus gp71. Thepanels are oriented with the high pH end to the left and the low pH end to the right. The region of the isoelectric dimension shown spans about pH 4 to 6.

encoded polypeptide chains was obtained using 2-deoxy-n- glucose to inhibit glycosylation of the MuLV glycoproteins (50). As shown by the L-[35S]methionine incorporation experiment in Fig. 4, the envelope glycoprotein precursors gPr89/91”“” (Lane A) were not formed by cells which had been preincubated with 2-deoxy-n-glucose (Lane B). Rather, these cells synthesize two smaller proteins with molecular weights of about 72,000 (Pr72”“) and 73,000 (Pr73”“) that were precipitated by the anti-gp70 antiserum. Furthermore, cells infected with the B4 clone of Eveline virus synthesize only Pr73”“” (data not shown) when incubated with 2-deoxy- n-glucose. These results suggest that Pr72en” and Pr73’“” are the unglycosylated polypeptide chains specified by the gp69 and gp71 genes, respectively. Their sizes indicate that Pr73”“” may contain about 8 amino acids more than Pr72”” and that both polypeptide chains contain approximately 600 amino acids.

Heterogeneous Sialylation of Envelope Glycoproteins--In addition to their size heterogeneity, the envelope glycoproteins are clearly heterogeneous in their isoelectric point (Fig. 3). To investigate the possibility that negative charge supplied by sialic acids might contribute to the “string-of-beads” appearance of the glycoprotein in a relatively acidic (pH 4 to 6) region of the isoelectric focusing dimension, an L-[““Slmethi- onine-labeled preparation of Eveline virus was treated with neuraminidase to remove the terminal sialic acid residues. Fig. 5 shows that the “string-of-beads” is eliminated by neuraminidase. Furthermore, a lightly labeled component in the control virions (Fig. 5A, arrow) is enriched following neuraminidase treatment (Fig. 5B, arrow). Because the asialoglycoprotein focuses at the top high pH end of the isoelectric focusing dimension, where the pH gradient rises very steeply

A B

#Jr gPr

-/Pr 73 -Pr 72

FIG. 4. Glycosylated and unglycosylated envelope glycoprotein precursors. Suspension cultures (5 ml) of Eveline cells (-2 x lO”/ml) were pulse-labeled with 40 pCi/ml of L-[35S]methionine for 30 min in methionine-free minimal essential medium. Glycosylation was in- hibited by preincubating (2 h) and labeling cells in medium containing 15 rnM 2-deoxy-n-glucose (Calbiochem) (50). The labeled cells were then lysed, the cell extracts were immune-precipitated by secondary antiserum precipitations using monospecific antibody to gp69/71 (40), and the immunoprecipitates were analyzed by electrophoresis in polyacrylamide gradient slab gels containing 0.1% sodium dodecyl sulfate. Lanes A and B contain the gp69/71specific precursors immune-precipitated from Eveline cells and electrophoresed in an 8 to 15% polyacrylamide gradient. Lane A is the glycosylated precursors gPr89/91’““. Lane B is the unglycosylated precursors Pr72/73’“‘, labeled in the presence of 2-deoxy-o-glucose.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

1344 Murine Leukemia Virus Glycoproteins

between pH 9 to 11, its isoelectric point homogeneity cannot be evaluated. Therefore, although this result shows that sialic acids occur in gp69/71, consistent with previous reports (55, 56), it cannot be concluded that the isoelectric point microheterogeneity is caused by heterogeneous sialylation.

Fig. 6. shows an analysis of n-[3H]glucosamine-labeled envelope glycoproteins from Eveline virus by “nonequilibrium pH gradient electrophoresis” (47), a method designed to maximize resolution of basic proteins. Desialylated gp69 and gp71 (Panel B) migrate in these gels as relatively homogeneous proteins compared with the heterogeneous mobility of the sialoglycoproteins (Panel A). However, the desialylated gp69 and gp71 retain some heterogeneity and the gp69 is slightly more basic than the gp71. As mentioned above, the relatively more acidic sialoglycoprotein components often appear to have slightly larger sizes than the more basic components. These results support the idea that heterogeneous sialylation is a major cause of gp69 and gp71 charge heterogeneity. Furthermore, heterogeneous sialylation may result in some size microheterogeneity within each glycoprotein family.

Oligosaccharides of the Envelope Glycoproteins-A DEAE-Sephadex chromatograph of tryptic peptides from D-

[“Hlglucosamine-labeled Eveline virus glycoproteins is shown in Fig. 7. The control glycopeptides chromatograph as a runoff peak followed by a heterodisperse spectrum of glucosamine- labeled material which binds to the column with variable affinity. Emerging from this broad spectrum is a component

IQP 6901

FIG. 5. Two-dimensional polyacrylamide gel electrophoresis of control and neuraminidase-treated Eveline virus gp69/71. Virion proteins were labeled with L-[%]methionine as described under “Exper- imental Procedures.” Purified virus was suspended in 10 mM Tris- HCl, 100 mM NaCl, 1 mM EDTA, pH 7.4, and was incubated with neuraminidase for 2 h at 37°C. Control samples were incubated in parallel without the enzyme. Control and neuraminidase-treated virus was denatured and electrophoresed in two dimensions essentially as described by O’Farrell (46). Only the gp69/71 region of the resulting autoradiograms is shown. Panel A is control Eveline virus. Panel B is neuraminidase-treated Eveline virus. The gels are oriented with the high pH end to the left and the low pH end to the right. The region of the isoelectric focusing dimension shown spans about pH 4 to pH 10. The arrows indicate the position of the asialoglycoprotein.

A

FIG. 6. Two-dimensional nonequilibrium pH gradient electrophoresis of control and neuraminidase-treated Eveline virus gp69/71. Virion proteins were labeled, purified, and neuraminidase-treated as in Fig. 5. Control and neuraminidase-treated virus was denatured and electrophoresed in the NEPHGE system as described by O’Farrell (47). Only the gp69/71 region of the resulting autoradiograms is shown. Panel A is control Eveline virus. Panel B is neuraminidase- treated Eveline virus. The gels are oriented with the high pH end to the left and the low pH end to the right.

Neurominidose

-Siolic acid

20 40 60 60 100 120

Fraction

FIG. 7. DEAE-Sephadex chromatography of tryptic glycopeptides of Eveline virus gp69/71. Virion proteins were labeled with D- [3H]glucosamine as described under “Experimental Procedures.” Neuraminidase and trypsin digestions and column chromatography procedures are described under “Experimental Procedures.” The column procedure was essentially that of Robertson et al. (49). The upper chromatogram (0) is the untreated control glycopeptides and the lower chromatogram (0) is the neuraminidase-treated glycopeptides. The NaCl gradient (0 to 0.3 M) was started at Fraction 20.

(Peak A) which is usually very prominent and which repro- ducibly elutes in Fractions 55 to 60. Neuraminidase treatment alters this profile. In this case, the runoff is closely followed by a peak (Fraction 20) of free sialic acid (49) which was released from the glycoprotein by neuraminidase and which cochromatographs with N-[‘%]acetylneuraminic acid. The broad heterogeneous spectrum of material which was present in the control column is eliminated by neuraminidase but Peak A is unaffected by the enzyme treatment. Additional neuraminidase treatments had no effect on the yield of this glycopeptide. These results imply that the peak A glycopeptide lacks sialic acid. Presumably, it binds to the column by an acidic amino acid in its peptide moiety. The broad spectrum


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


also contains a heterogeneous group of sialoglycopeptides and desialylation of these glycopeptides apparently causes them to elute in the column runoff peak.

The DEAE-Sephadex chromatography results shown above suggest that the sialoglycopeptides of gp69/71 are very heterogeneous. However, those results provide no information concerning the number of different sialylated carbohydrate units in the protein or the number of neutral glycopeptides which elute in the DEAE-Sephadex column runoff peak. In addition, they do not indicate whether the neutral glycopeptides in the runoff are merely “asialo” forms of the sialoglycopeptides. To address these questions we collected the asialoglycopeptides which elute in the DEAE-Sephadex runoff fraction (Fig. 7), digested them with pronase, and analyzed their oligosaccharides by gel filtration on a Sephadex G-50 column (Fig. 8). It can be seen that the DEAE-Sephadex column runoff of the control untreated glycoprotein contained only one prominent oligosaccharide peak (40 to 47 ml) which appeared to have partially separated into two components, whereas the DEAE-Sephadex runoff glycopeptides from the neuraminidase-treated glycoprotein contained an additional heterogeneous group of larger oligosaccharides. Sialylated oligosaccharides in proteins are commonly large and complex in structure compared with many of the oligosaccharides which contain only neutral sugars (57-59).

These two different classes of oligosaccharides which are distinguished both by their size and by the presence of sialic acid (Figs. 7 and 8) probably correspond in general structure to the N-glycosidically linked oligosaccharides which have been designated as “complex type “ and “high mannose type” (57-59), although chemical analysis will obviously be neces- sary to establish their structures. After incubation of D- [3H]glucosamine-labeled envelope glycoprotein with alkali (0.1 N NaOH at 25°C for 24 h) in conditions that cause cleavage of 0-glycosidic bonds to serine or threonine (51),

25 35 45 55 65

Volume (ml) FIG. 8. Gel filtration of Eveline virus control and neuraminidase-

treated glycopeptides not adsorbed to the DEAE-Sephadex columns. The [3H]glucosamine-labeled runoff material from the DEAE-Seph- adex columns (Fig. 7) was pooled and was pronase-digested as described under “Experimental Procedures.” 0, control glycopeptides unadsorbed to DEAE-Sephadex column; 0, neuraminidase-treated glycopeptides unadsorbed to DEAE-Sephadex column.

most of the radioactivity in the reneutralized samples eluted from a Sephadex G-50 column in the excluded volume (data not shown), indicating that it had not been cleaved from the glycoprotein. This observation is consistent with at least a large fraction of the labeled sugars being present in aspara- gine-linked carbohydrate units.

Fig. 9 shows an analysis of n-[3H]glucosamine-labeled pronase digested glycopeptides from Eveline virus using high voltage electrophoresis. In this experiment the free [3H]sialic acid is not seen because it has electrophoresed off the paper toward the anode. The neutral asialoglycopeptides migrate in these conditions near the electroosmotic front as a rather homogeneous peak in Fractions 0 to 5, whereas the sialoglycopeptides migrate heterogeneously toward the anode. These results support the conclusion that the envelope glycoproteins contain neutral carbohydrate(s) as well as a heterogeneous group of sialylated oligosaccharides. Additional pronase digestion did not alter the results.

In the experiments described above, a large proportion of the n-[3H]glucosamine label is released from the glycoprotein by neuraminidase treatment. This was expected because D- [3H]glucosamine is generally incorporated into glycoproteins as either N-acetylglucosamine or as sialic acid (60). We have determined that approximately 45 to 60% of the radioactivity present in n-[3H]glucosamine labeled virion envelope glycoproteins is converted by neuraminidase treatment into a form that is soluble in cold 5% trichloroacetic acid.

The DEAE-Sephadex column chromatography experiment was repeated using the B4 clone of Eveline virus. Although the amount of radioactivity was lower due to the lower yield of virus and the inefficiency of labeling the infected SC-1 cells with nj3H]glucosamine (see “Experimental Procedures”), the results were otherwise the same as the whole Eveline virus. In particular, all of the glycopeptides observed in the whole Eveline virus glycoproteins were also present in the gp71 of the B4 virus (data not shown).

Eveline MuLV Is Not Identical to Friend MuLV-The virus produced by Eveline cells is generally considered to be Friend MuLV (34, 36-39). However, during the course of this work and previous studies (40,61), we have found that Eveline MuLV differs significantly from Friend MuLV (F-MuLV). Fig. 10 shows the results of an experiment in which Eveline cells

1

60 - 0 Control l Neurommldass

-9 ” 4 8 12 16 20 24 26 32 36 40 Fraction (cm)

FIG. 9. High voltage paper electrophoresis of pronase glycopeptides of Eveline virus gp69/71. Virion proteins were labeled with D- [JH]glucosamine, treated with neuraminidase and digested with pronase, and the glycopeptides were electrophoresed as described under “Experimental Procedures.” 0, control sample; 0, neuraminidase- treated sample. The figure is oriented with the anode to the left.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


FIG. 10. Cell surface glycoproteins of Eveline cells and Friend erythroleukemia cells, as revealed by lactoperoxidase-catalyzed iodi- nation of cells with [‘251]iodine. The erythroleukemia cells are the F4- 6/K cell line of Ostertag et al. (44). After removal of unreacted iodine, the cellular proteins were solubilized, immunoprecipitated with monospecific antibody to gp69/71, and analyzed by electrophoresis as described under “Experimental Procedures.” The figure shows an autoradiogram of the polyacrylamide gel. A longer exposure of the fiim revealed the clear labeling of both gp69 and gp71 in the Eveline cell sample.

and Friend erythroleukemia cells (cell line F4-6/K) were radioiodinated on their plasma membranes using lactoperoxidase. The corresponding radioactive glycoproteins were precipitated from the cell extracts with antiserum to gp69/71 and the immunoprecipitates were analyzed by electrophoresis in polyacrylamide gels in the presence of sodium dodecyl sulfate. The Friend cell virion glycoprotein (gp75) is larger than the glycoprotein on Eveline cells (gp69/71). Furthermore, other experiments indicate more clearly that Eveline cells contain both gp69 and gp71 on their surface, whereas cells infected with F-MuLV contain only one surface glycoprotein. Simi- larly, some of the p30 precursors of the two viruses have different molecular weights (40, 61). In addition, the virus proteins produced by Friend erythroleukemia cells were found to be indistinguishable from those encoded by the original strain of Friend virus. The distinctive properties of F-MuLV and Eveline MuLV were also independent of the cell line in which the viruses were grown. We suggest that Eveline MuLV may be a variant of Friend MuLV and that their precise relationship will require additional information.

DISCUSSION

General Comments-Our results demonstrate that the presence of two size classes of the MuLV envelope glycoprotein from Eveline cells is the result of two different viral genomes in the culture rather than the result of heterogeneous processing (i.e. proteolysis or glycosylation) of a single polypeptide chain. When cloned virus which contains only one glycoprotein gene is employed, only a single size class of envelope glycoproteins is produced. Furthermore, studies employing 2- deoxy-D-glucose to inhibit glycosylation support the conclusion that the two envelope glycoprotein genes present in uncloned Eveline MuLV encode for polypeptide chains which

have an apparent size difference of approximately 1,000 dal- tons. In addition to these molecular genetic studies, our results suggest that the glycoprotein charge heterogeneity is caused principally by heterogeneous sialylation of its oligosaccharides.

Size Heterogeneity-We have analyzed the envelope glycoproteins encoded by four different virus clones derived from the Eveline virus. Two of these cloned viruses encode only for the synthesis of gp71, whereas the other two encode for the synthesis of both gp69 and gp71. The latter two clones conceivably could have been caused by infection with virus aggregates present in the preparation of Eveline virus. However, we did not see virus aggregates in the electron microscope. Alternatively, such clones would be expected if the RNA genome which codes for gp69 can combine together with the genome encoding for gp71 to form heterozygous MuLV particles. Evidence for heterozygotes was previously obtained for Rous sarcoma virus (62) and for Moloney MuLV (63). Presum- ing that the MuLV genome is diploid and that the gp69 and gp71 genomes are equally prevalent, we would expect half of the virus clones to be heterozygotes and each type of homo- zygote to occur one-fomth of the time. We have initiated a test of these matters and a search for homozygous gp69 virus by isolating a large number of clones from Eveline virus particles purified by sucrose gradient sedimentation (63). Availability of these clonally purified homozygous MuLVs should also facilitate a more detailed analysis of the structures of gp69 and gp71.

Charge Heterogeneity-Although charge heterogeneity of the envelope glycoproteins has previously been observed and ascribed to sialylation (55, 56), the evidence obtained was inconclusive. Witte et al. (55) examined the gp69/71 of Mo- loney-sarcoma-leukemia virus by two-dimensional gels, similar to those used in Figs. 3 and 5. They observed an “acidic/heterogeneous” band which was not resolved into multiple components. This material was more homogeneous “near the limits of definition of our tube gel isoelectric focusing system” after treatment with neuraminidase. Their conclusion about the homogeneity of a protein in this region of the isoelectric focusing gel suffers from the same deficiencies as our gel. That is, there is little resolution in this region of the system. The control gp69/71 focused in their experiment as a heterodisperse band that was approximately twice as broad as the neuraminidase treated material which focused at the top of the isoelectric focusing gel. Although the untreated glycoprotein appeared in the isoelectric focusing dimension as a broad band, other viral proteins on their gel appeared to be similarly broad. It is therefore unclear to us whether they observed true heterogeneity or incomplete focusing. Mar- quardt et al. (56) electrophoresed isolated gp70 at two pH values (4.3 and 8.9) in nondenaturing buffer. They described a broadly migrating gp70 band which was made more homog- enous after treatment with neuraminidase. However, the analysis of untreated control glycoprotein was not shown and the electrophoretic position of the broad band was not indicated. Furthermore, approximately half of the neuraminidase- treated material remained as an aggregate at the interface between the stacking and running gels. Our two-dimensional NEPHGE separations (Fig. 6) in denaturing conditions show that neuraminidase treatment causes the MuLV glycoprotein to migrate in a more basic and relatively homogeneous fash- ion. This conclusion seems unambiguous because the NEPHGE method was designed to accentuate the resolution of basic proteins. Even under these conditions, the more basic asialoglycoprotein spots appear relatively homogeneous.

Our oligosaccharide analyses are consistent with the above conclusions and suggest that the sialoglycopeptides of MuLV


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


envelope glycoproteins are indeed highly heterogeneous. For example, when analyzed by DEAE-Sephadex chromatography, the tryptic sialoglycopeptides are partially resolved into numerous minor components (Fig. 7). Similar heterogeneity was also observed by high voltage electrophoresis of pronase derived glycopeptides (Fig. 9). Moreover, the uncharged oligosaccharides which are produced from the sialoglycopeptides by neuraminidase treatment are larger and more heterogeneous in size than the neutral oligosaccharides which are present in the control untreated glycoprotein (Fig. 8). Based on these results, we propose that each glycoprotein molecule probably contains several large complex oligosaccharide units which are heterogeneously sialylated. We should mention that the complex oligosaccharides in proteins generally contain no more than 3 or 4 sialic acid residues (57-59) and that heterogeneous sialylation of glycoproteins (64-67) and oligosaccharide units (49, 58) has previously been observed. For example, the carbohydrate side chains of vesicular stomatitis virus glycoproteins are reported to contain either 0, 1, 2, or 3 sialic acid residues (49). Additionally, it should be noted that complex oligosaccharide units may also be structurally heterogeneous in their neutral sugar composition (57-59).

The envelope glycoproteins also appear to contain at least two different neutral oligosaccharides which are smaller than the sialylated oligosaccharides. One type of neutral carbohydrate occurs in a tryptic glycopeptide that adsorbs strongly to DEAE-Sephadex (Peak A in Fig. 7). Presumably, its binding is caused by an acidic amino acid in its peptide moiety. In addition, there is at least one neutral tryptic glycopeptide which elutes in the DEAE-Sephadex column runoff. This latter glycopeptide, after treatment with pronase, migrates on Sephadex G-50 as a single peak which appears to be partially resolved into two components (Fig. 8).

These results indicate that the MuLV envelope glycoproteins are glycosylated at several different sites with different types of oligosaccharide units. Presumably the two different classes of oligosaccharides observed in these experiments may correspond in general structure to the types of N-glycosidi- tally linked oligosaccharide units which are commonly found in glycoproteins and have been described by others (64-67). The alkaline stability of their linkage to the protein is consistent with this suggestion. Furthermore, Trowbridge et al. (68) have also recently shown that the pronase digested glycopeptides of MuLV envelope glycoprotein can be separated into two different sized components by gel filtration, and they found that only the smaller size class is susceptible to digestion by a mixture of a-mannosidase and N-acetyl-P-glucosamini- dase. This result supported their hypothesis that the smaller oligosaccharide(s) may belong to the high mannose (simple) type which have been often observed in glycoproteins.

Generality of Results Concerning gp69/71 Size Heteroge- neity-A major question raised by our cloning results concerns their generality. In other words, are all cell lines which produce two different sized MuLV glycoproteins simultaneously infected with at least two different viral genomes? Although a definitive answer to this question will obviously require cloning experiments with virus from other cell lines, there has been no published evidence that cells infected with a single MuLV genome produce more than one size class of glycoprotein. Even in cases where cloned virus may have been used to establish a cultured cell line, genetic divergence of multiple viral genomes within the culture cannot be excluded. Also, as mentioned above, cloned virus particles may be heterozygotes since virions contain at least two 35 S RNA subunits. Unless evidence is presented to the contrary, it therefore seems reasonable to believe that each haploid MuLV genome has only one enu gene (69) which produces only one size class of

envelope glycoprotein. Our results also imply that the type of cell does not significantly influence the size of the fully processed glycoprotein (Fig. 2). We have also obtained evidence supporting these conclusions using Rauscher MuLV. Al- though uncloned Rauscher virus contains the gp69/71 doublet (29,30,32), we have recently studied a clone of Rauscher virus which encodes only gp71.”

Origin of the Multiple Viral Genomes-Based on these considerations, it would appear that many of the established cell lines commonly used as sources of MuLV and of other mammalian C-type retroviruses may contain two or more viral genomes which encode for different size classes of envelope glycoprotein. It is therefore important to consider the origin of these multiple viral genomes.

Various studies suggest that the discrete size classes of glycoproteins which commonly occur in virus preparations are closely related variants. For example, peptide mapping studies and serological studies of gp69 and gp71 of Moloney and Rauscher MuLVs have indicated that the two glycoproteins are indistinguishable (13, 32). Furthermore, the gp69/71 of Eveline cells behaves as a single material when analyzed by immunological methods (37-39). In addition, we did not detect any tryptic glycopeptides in the uncloned Eveline virus which were absent in the cloned B4 virus. The simplest interpreta- tion of these results is that the two glycoproteins are highly homologous and are encoded by closely related MuLV genomes. This latter idea is consistent with our two-dimensional separations of proteins from Eveline virus and from the cloned B4 virus; the only proteins of these virus preparations which differed significantly were the envelope glycoproteins. One possible explanation consistent with our results is that a small deletion (sufficient to code for about 8 amino acids) in the gp71 gene produces a viral genome which encodes for gp69.

It would therefore appear that minor variants with altered envelope glycoprotein genes may occur frequently in preparations of mammalian C-type retroviruses. Conceivably, such glycoprotein variants may arise relatively frequently because of some special property of this region of the genome. Alter- natively, such variants may frequently occur because they are selected for, during virus replication in Go. For example, such variants in the envelope glycoprotein might permit the virus and the tumor cells to escape the immune response of the host or they might facilitate infection of tissues which would otherwise be resistant.

Acknowledgments-We are indebted to L. H. Evans and S. L. Dresler for their stimulating participation in early phases of this work, and to Ethel Polonoff for excellent technical assistance.

1. 2.

3.

4.

5. 6. 7.

8.

9. 10.

11.

REFERENCES

McLellan, W., and August, J. (1976) J. Viral. 20, 627-636 Witte, O., Weissman, I., and Kaplan, H. (1973) Proc. Natl. Acad.

Sci. U. S. A. ‘70, 36-40 Kennel, S., Del Villano, B., Levy, R., and Lerner, R. (1973)

Virology 55, 464-475 Ikeda, H., Pincus, T., Yoshiki, T., Strand, M., August, J., Boyse,

E., and Mellors, R. (1974) J. Viral. 14, 1274-1280 Delarco, J., and Todaro, G. (1976) Cell 8, 365-371 Besmer, P., and Baltimore, D. (1977) J. Virol. 21, 965-973 Ishimoto, A., Hartley, J., and Rowe, W. (1977) Virology 81,

263-269 Van Zaane, D., Dekker-Michielsen, M., and Bloemers, H. (1976)

Virology 75, 113-129 Kennel, S., and Feldman, J. (1976) Cancer Res. 36, 200-208 Ledbetter, J., Nowinski, R., and Emery, S. (1977) J. Viral. 22,

65-73 Bolognesi, D. P., Montelaro, R. C., Frank, H., and Schafer, W.

(1978) Science 199, 183-186

’ Unpublished observations.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


12. Witte, O., and Baltimore, D. (1977) Cell 11, 505-511 13. Elder, J., Jensen, F., Bryant, M., and Lerner, R. (1977) Nature

267,23-28 14. Strand, M., and August, J. (1976) Virology 75, 130-144 15. Tung, J., Shen, F., Fleissner, E., and Boyse, E. (1976) J. Exp.

Med. 143,969-974 16. Tung, J., Fleissner, E., Vitetta, E., and Boyse, E. (1975) J. Exp.

Med.142,518-523 17. Obata, Y., Ikeda, H., Stockert, E., and Boyse, E. (1975) J. Exp.

Med. 141, 188-197 18. Strand, M., August, J., and Jaenisch, R. (1977) Virology 76,

886-890 19. Lerner, R., Wilson, C., Del Villano, B., McConahey, P., and Dixon,

F. (1976) J. Exp. Med. 143, 151-166 20. Del Villano, B., and Lerner, R. (1976) Nature 259,497-499 21. Ihle, J., Denny, T., and Bolognesi, D. (1976) J. Virol. 17, 727-736 22. Ihle, J., Hanna, M., Roberson, L., and Kenney, F. (1974) J. Exp.

Med. 139, 1568-1581 23. Ihle, J., and Hanna, M. (1977) Contemp. Topics Immunohiol. 6,

169-194 24. Yoshiki, T., Mellors, R., Strand, M., and August, J. (1974) J. Exp.

Med. 140, 1011-1027 25. Troxler, D., Lowy, D., Howk, R., Young, H., and Scolnick, E.

(1977) Proc. Natl. Acad. Sci. U. S. A. 74,4671-4675 26. Elder, J., Gautsch, J., Jensen, F., Lerner, R., Hartley, J., and

Rowe, W. (1977) Proc. Natl. Acad. Sci. U. S. A. 74,4676-4680 27. Schafer, W., Schwarz, H., Theil, H.-J., Fischinger, P. J., and

Bolognesi, D. P. (1977) Virology 83, 207-210 28. Strand, M., Bilello, J., Shapiro, S., and August, J. (1976) in Tumor

Virus Infections and Immunity (Crowell, R., Friedman, H., and Prier, J., eds) pp. l-34, University Park Press, Baltimore

29. Strand, M., and August, J. (1973) J. Biol. Chem. 248, 5627-5633 30. Naso, R., Arcement, L., Karshin, W., Jamjoom, G., and Arling-

haus, R. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 2326-2330 31. August, J., Bolognesi, D., Fleissner, E., Gilden, I-t., and Nowinski,

K. (1974) Virology 60, 595-601 32. Krantz, M. J., Strand, M., and August, J. T. (1977) J. Virol. 22,

804-815 33. Karshin, W. L., Arcement, L. J., Naso, R. B., and Arlinghaus, R.

B. (1977) J. Virol. 23, 787-798 34. Seifert, E., Claviez, M., Frank, H., Hunsmann, G., Schwarz, H.,

and Schafer, W. (1975) 2. Naturforsch. Teil. B 30, 698-700 35. Luftig, R., McMillan, P., and Bolognesi, D. (1974) Cancer Res.

34,3303-3310 36. Moennig, V., Frank, H., Hunsmann, G., Schneider, I., and Schafer,

W. (1974) Virology 61, 100-111 37. Bolognesi, D. B., Collins, J. J., Leis, J. P., Moennig, V., Schafer,

W., and Atkinson, P. H. (1975) J. Virol. 16, 1453-1463 38. Schafer, W., Fischinger, P. J., Collins, J. J., and Bolognesi, D. P.

(1977) J. Virol. 21, 35-40 39. Collins, J., Bolio, L., Denny, T., and Bolognesi, D. (1977) J. Viral.

21, 113-118

40. Evans, L., Dresler, S., and Kabat, D. (1977) J. Viral. 24, 865-874 41. Hartley, J., and Rowe, W. (1975) Virology 65, 128-134 42. Bassin, R., Tuttle, N., and Fischinger, P. (1971) Nature 229,

564-566 43. Sherton, C., Evans, L., Polonoff, E., and Kabat, D. (1976) J. Virol.

19, 118-125 44. Ostertag, W., Melderis, H., Steinheider, G., Kluge, N., and Dube,

S. (1972) Nature New Biol. 239, 231-234 45. Vitetta, E., Baur, S., and Uhr, J. (1971) J. Exp. Med. 134,242-264 46. O’Farrell, P. (1975) J. Biol. Chem. 250,4007-4021 47. O’Farrell, P. Z., Goodman, H., and O’Farrell, P. H. (1977) Cell 12,

1133-1142 48. Bonner, W., and Laskey, R. (1974) Eur. J. Biochem. 46,83-88 49. Robertson, J., Etchison, J., and Summers, D. (1976) J. Viral. 19,

871-878 50. Shapiro, S., Strand, M., and August, J. (1976) J. Mol. Biol. 107,

459-477 51. Sharon, N. (1975) Complex Carbohydrate, Addison-Wesley,

Reading, Mass. 52. Kiely, M., McKnight, G. S., and Schimke, R. T. (1976) J. Biol.

Chem. 251,5490-5495 53. Rothman, J., and Lodish, H. (1977) Nature 269, 775-780 54. Robbins, P. W., Hubbard, S. C., Turco, S. J., and Wirth, D. F.

(1977) Cell 12, 893-900 55. Witte, O., Tsukamoto-Adey, A., and Weissman, I. (1977) Virology

76,539-553 56. Marquardt, H., Gilden, R., and Oroszlan, S. (1977) Biochemistry

16,710-717 57. Kornfeld, R., and Kornfeld, S. (1976) Annu. Reu. Biochem. 45,

217-237 58. Jarnefelt, J., Finne, J., Krusius, T., and Rauvala, H. (1978) Trends

Biochem. Sci. 3, 110-114 59. Baenziaer. J.. and Kornfeld. S. (1974) J. Biol. Chem. 249.

727017281 60. Knopf, P. Sasso, E., Destree, A., and Melchers, F. (1975) Bio-

chemistry 14,4136-4143 61. Sherton, C., Dresler, S., Polonoff, E., Webb, M., and Kabat, D.

(1979) Cancer Res. 38, 1426-1433 62. Wyke, J. (1975) Biochim. Biophys. Acta 417,91-121 63. McCarter, J. (1977) J. Virol. 22, 9-15 64. Gati, W., and Straub, P. W. (1978) J. Biol. Chem. 253, 1315-1321 65. Tate. S.. and Ross. M. (1977) J. Biol. Chem. 252. 6042-6045 66. Lusis, A., Breen, G., and Paigen, K. (1977) J. Biol. Chem. 252,

4613-4618 67. Zimmerman, E., and Wilson, J. (1976) in Onto-Deuelopmental

Gene Expression (Fishman, W. W., and Sell, S., eds) pp. 139-145, Academic Press, New York

68. Trowbridge, I. S., Hyman, R., and Hazauskas, C. (1978) Cell 14, 21-32

69. Baltimore, D. (1974) Cold Spring Harbor Symp. Quant. Biol. 39, 1187-1200


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

M J Murray and D Kabatmembrane envelope glycoproteins gp69/71.

Genetic and sialylation sources of heterogeneity of the murine leukemia virus

1979, 254:1340-1348.J. Biol. Chem.

http://www.jbc.org/content/254/4/1340.citation

Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/254/4/1340.citation.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/content/254/4/1340.citation

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;254/4/1340&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/254/4/1340.citation

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=254/4/1340&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/254/4/1340.citation

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/254/4/1340.citation.full.html#ref-list-1

http://www.jbc.org/

Genetic and Siaiylation Sources of Heterogeneity of the Murine ...

Documents