Structural Proteins of Oncogenic Ribonucleic Acid Viruses

Two Rauscher murine leukemia virus polypeptides with apparent molecular weights of 69,000 and 71,000 have been purified and characterized. These polypeptides contained a previously undescribed interspecies antigenic determinant of mammalian oncogenic C type RNA viruses (interspec II), as was demonstrated by the reaction of the murine antigen with anti-feline leukemia virus serum. Radioimmunoassay analysis showed that both polypeptides were precipitated by the anti-feline virus serum, indicating that they each contained the interspecies determinant or that they were closely associated and coprecipitated. The interspec II antigen was distinguished from the known gs-3 interspecies antigen (interspec I) and the virus RNA-dependent DNA polymerase by protein purification, physical properties, and immunological analysis. The new antigen was not detected by competition radioimmunoassay in uninfected mouse cells or mouse cells productively infected with vesicular stomatitis virus.

These polypeptides contained a previously undescribed interspecies antigenic determinant of mammalian oncogenic C type RNA viruses (interspec II), as was demonstrated by the reaction of the murine antigen with anti-feline leukemia virus serum. Radioimmunoassay analysis showed that both polypeptides were precipitated by the anti-feline virus serum, indicating that they each contained the interspecies determinant or that they were closely associated and coprecipitated.
The interspec II antigen was distinguished from the known gs-3 interspecies antigen (interspec I) and the virus RNA-dependent DNA polymerase by protein purification, physical properties, and immunological analysis.
The new antigen was not detected by competition radioimmunoassay in uninfected mouse cells or mouse cells productively infected with vesicular stomatitis virus.
The structural proteins of oncogenic C type RNA viruses have proven to be valuable reagents for analysis of virus gene expression and for characterization of viral particles.
Recently, these proteins have become increasingly important in the search for possible oncogenie viruses in human neoplasia.
Some of the most useful of these proteins are those containing antigenic determinants found in other viruses as well (group-specific (gs) antigens).
Two classes of such antigens are known: (a) those that are found only in viruses of the same animal species (speciesspecific antigens), and (6) those that are present in viruses from animals of different species (interspecies antigens).
Recently it was found that the RNA-dependent DNA polymerases of mammalian C type viruses also contain interspecies antigenic determinants.
Antiserum against the polymerase of either murine virus or feline virus inhibited the DNA polymerase of mouse, rat, cat, or hamster viruses (13,14).
We now report the isolation and characterization of other major Rauscher murine leukemia virus polypeptides containing an interspecies antigen.
Until a standard nomenclature for the proteins of C type viruses is developed, we shall refer to this antigen as interspec II.

Purification of Virus Proteins-Frozen
Rauscher MuLV,i 92 mg of total protein in 50 ml of 50 InM sodium citrate, was thawed and dialyzed against 2 liters of 5 InM Tris-HCl, pH 7.6, and 1 mM EDTA for 4 hours with one change of dialysate.
The virus was then centrifuged at 108,000 X g for 2 hours, and the pellet, 60 mg of protein, suspended in 20 ml of 5 mM Tris-HCl, pH 9.2, 1 mM EDTA, and 2 M KCl.
This virus suspension was sonicated for a total of 90 s in a Branson Sonifier, and then centrifuged at 108,000 X g for 2 hours. The supernatant, 18 mg of protein, was dialyzed for 15 hours against 2 liters of 10 IDM N ,N-bis(Bhydroxyethyl)-2-aminoethanesulfonic acid pH 6.5, and 1 InM EDTA, with one change of dialysate, and applied to a column (1.5 x 10 cm) of phosphocellulose (Whatman P-11) previously equilibrated with the same buffer. The column was washed with the starting buffer and then eluted at a rate of 10 ml per hour, with a total of 200 ml of buffer solution containing 10 mM BES, pH 6.5, 1 mM EDTA and a linear gradient of 0 to 1.0 M KCI.
Fractions of 2 ml were collected. Following the flow through of unabsorbed proteins, two major protein fractions were eluted.
The first (phosphocellulose Fraction I) contained 6.2 mg of protein, and eluted at 2.5 to 15 InM KCl.
The second phosphocellulose fraction contained 3.9 mg of protein and eluted at 20 to 27 mM KCl.
Both of these fractions contained protein which was precipitated by anti-feline leukemia virus serum as well as with anti-MuLV serum. Phosphocellulose Fraction I was found to contain two polypeptides (see "Results"), which were not further purified.
The protein of the second phosphocellulose fraction was precipitated by the addition of (NH&SO4 to 75yo saturation aud collected by centrifugation. The precipitate was suspended in 10 m&l UES, pH 6.5, and 1.0 M SaCI, dialyzed, and applied to a column of Sephadex G-200 (1.5 x 50 cm) equilibrated in the same buffer. Fractions of 2.0 ml were collected at a flow rate of 10 ml per hour.
The major peak (Sephadex fraction), 2 mg of total protein, contained an ap parently homogeneous protein of 30,000 daltons (see "Results").
Purification of the Rickard FeLV 27,000-dalton structural protein containing the feline interspec I determinant was carried out in the same manner by phosphocellulose and Sephadex column chromatography.
Antisera-Anti-Rauscher MuLV serum and anti-Rickard FeLV serum were obtained from rabbits injected with purified virus particles that had been degraded by incubation at 45" foi 1 min with 0.2% sodium dodecyl sulfate or 0.4y0 Triton X-100 and mixed with an equal volume of Freund's complete adjuvant.
In some experiments, as indicated, the immunoglobulin was partially purified by Na&Oe precipitation as described by Kekwick (15). Rabbit and goat antisera against the purified interspee I and interspec II antigens were prepared by multiple injections of purified proteins in Freund's comp1et.  (18), and Oroszlan et al. (19). The reaction mixture contained the following: 0.03 ml of normal rabbit serum diluted 6-fold in a buffer solution containing 20 InM Tris-HCl, pH 7.6, 100 m&f NaCl, 1 mM EDTA, and 2 mg per ml of Fraction V bovine serum albumin, 0.01 ml of 'ZjI-labeled virus protein (2 to 5 ng of protein containing lo4 to 10" cpm per ng, as indicated), and 0.01 ml of diluted antiserum, as indicated.
All antigen and antibody proteins were diluted in a buffer solution containing 20 m&r Tris-HCl, pH 7.6, 100 rnnl NaCl, 1 mM EDTA, and 20 mg per ml of crystalline bovine serum albumin.
The reaction mixture was incubated at 37" for 3 hours, after which 0.03 ml of horse anti-rabbit IgG was added to precipitate the antigenantibody complex.
The mixture was incubated for 12 hours at 2-4".
Cold buffer solution (0.5 ml) containing 20 mM Tris-HCI, pII 7.6, 100 mM NaCl, and 1 InM EDTA was added, and the ljrecipitate was collected by centrifugation at 4". The pellet was washed twice with the same buffer solution, and the r2jI-labeled antigen present in the precipitate was measured in a gamma counter.
The same reaction mixture was used for the competition radioimmunoassay.
Competing proteins were diluted in a buffer solution containing 20 mM Tris-HCl, pH 7.6, 100 mM NaCl, 1 mM EDTA, and 20 mg per ml of crystalline bovine serum albumin, and were added in 0.01 ml immediately before the antiserum. When virus particles or cells were tested, Triton X-100, 0.4% final concentration, was added to the virus or cell preparation, and the suspension was incubated at 37" for 10 min. In this case, the buffer solution used for serial dilution of the fractions also contained 0.4y0 Triton X-100.
After incubation at 37" for 30 min the reactions were terminated by the addition of 3.0 ml of cold 5y0 trichloracetic acid containing 20 mM NAPPi.
The acid-insoluble radioactivity was collected on Whatman GF/C glass fiber discs, washed with the trichloracetic acid-PP i solution, and measured by liquid scintillation spectroscopy.
In experiments testing the effect of rabbit antipolymerasc antibodies, the purified antiserum immunoglobulin and equivalent control immunoglobulin were added to complete reaction mixtures prior to incubation.
The polymerase activity present with 3 to 4 pg of purified virus was 50% inhibited by 15 to 20 pg of immunoglobulin protein.
Proteins competing with the polymerase for antibody binding were tested in the same manner and were added immediately prior to the antiserum.
Viruses-Rauscher murine leukemia virus was kindly provided by Drs. S. Mayyasi and D. Larson of t.he John L. Smith Memorial for Cancer Research.
The virus was propagated in a RALB/c mouse bone marrow culture (JLS-V9) continuously infected with Rauscher virus as described by Wright et al. (20), harvested, and purified as described previously (21), and suspended in 50 mM sodium citrate.
A rat tropic Kirsten strain of murine sarcoma virus, propagated in a clone of productively infected, transformed normal rat kidney cells (22) Avian myeloblastosis virus in plasma from infected chickens was generously supplied by Dr. J. Beard, Life Sciences, Inc. The virus was concentrated by centrifugation onto 60% sucrose and purified by double density gradient centrifugation in sucrose. Herpes simplex virus (subtype 1), grown in human epidermoid carcinoma No. 2 cells (HEp-2), and purified as described by Spear and Roizman (32), was a gift from Dr. B. Roizman, University of Chicago.
Vesicular stomatitis virus grown in HeLa cells (33) was a gift from Dr. D. E. Summers, Albert Einstein College of Medicine.
Mycoplasma arginine was donated by Dr. R. Steeves.

RESULTS
Protein Purity and iMolecular Weight--The homogeneity and molecular weight of the purified murine virus proteins were analyzed by electrophoresis in polyacrylamide gels containing sodium dodecyl sulfate.
The phosphocellulose Fraction I contained two polypeptides with apparent molecular weights of 69,000 and 71,000 (Fig. 1C). These polypeptides correspond to major components of the crude virus and, as shown below, were found to comprise 5 to 10% of the protein of undegraded virus and to contain the interspec II antigen.
In recent experiments it was found that this purified fraction also contains small amounts of sialic acid, glucosamine, galactose, and possibly fucose (data not shown), suggesting that these polypeptides comprise glycoprotein components of the virus.
In this case the actual molecular weight of the molecules may be lower as the migration of glycoproteins in sodium dodecyl sulfate polyacrylamide gel electrophoresis is not an accurate measurement of size.
The Sephadex fraction contained an apparently homogeneous component of molecular weight 30,000 (Fig. ID).
This protein is the major structural component of the virus, and is known to contain interspec I, the previously reported interspecies antigenic determinant (gs-3 or gs-interspec) (l-10).

Gel Difusion
Evidence of Two Interspecies Antigens-Each of the fractions purified from Rauscher MuLV contained protein which reacted with anti-FeLV serum as well as with anti-MuLV serum, indicating that both murine virus proteins contained antigens common to feline virus (Fig. 2). Moreover, the precipitin patterns of nonidentity indicated that the antigenic determinants were different. This was true for all of the determinants bound by the antimurine antibodies as well as the interspecies determinants bound by the antifeline antibodies. The 2. Gel diffusion analysis of interspec I and interspec II. Immunodiffusion plates were prepared using 75 X 35-mm microscope slides coated with 4.5 ml of 0.8y0 agarose in 50 mM Tris, pH 7.2, 100 mM NaCI, and 0.5 M glycine, and wells of 25.~1 volume were cut immediately prior to use. The rabbit antisera had been concentrated approximately 8-fold by Na&O diffuseness of the precipitin liue of interspec II was commonly observed aud was often seen to resolve into two distinct bands; this is possibly related to the separation of the two polypeptides of this protein fraction.
There appeared to be partial merging between the precipitin of interspec I and interspec II with antifeline serum but we believe this is an artifact due to nouspecific absorption of autiserum by the proteins or to adherence of the proteins to one another.
As showu below, antiserum against each of the purified interspec antigens was monospecific even in the highly sensitive radioimmunoassay.
Radioimmunoassay of Interspec 11-111 order to determiue whether one or both of the polypeptides cont.ained the interspec II antigen, binding of the proteins by anti-MuLV serum aud anti-FeLV serum was quantitatively measured by the radioimmunoassay.
All of the acid-precipitable radioactivity of the 12SI-labeled interspec II antigen was precipitated by either ;\luLV or FeLV antiserum (Fig. 3). Autoradiography of the 69,000aud 71,000.dalton polypeptides separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis showed that both were equally labeled (data not shown).
These results indicate that both polypeptides contain the interspee II autigenic determinants or that they were closely associated and coprecipitated.
Other radioimmuuoassay studies to be reported elsewhere show that the 69,000 to 71,000 polypeptides coutain speciesspecific aud possibly type-specific antigens as well as the inferspec II determinants.
The presence of both species-specific and interspecies antigens has beeu demonstrated as well with the major structural protein containing the interspec I determinant (18,35,36).
Co?nparison of fnterspec I and II by Competition Radioim-??zunoassay-hdditioual evidence that the interspecics antigenic determinauts of interspec II differed from those of interspec I was obtaiued by use of the competitiou radioimmuuoassay. In these esperimellts competition between the interspecies determinants of the two murine proteins for bind& by anti-FcLV serum was tested with both IT-interspec I and ]2"I-interspec II antigens (Fig. 4). l&ding of XIuLV '*"I-interspec I antigen by anti-FeLV serum was not blocked by addition of interspec II; nor was binding of the ;\4uLV 12Gherspec IT antigen by anti-FeLV serum bloctked by FeLV interspec 1. Ill COlltrol csperiments binding of the labeled proteins was blocked by uulabelcd counterparts, as expected. Thus, 110 autigeuic crossreactivity could be demonstrated between the interspecies determinants of the two proteins by this highly sensitive assay. The assays were performed as described under "Methods" with 2.5 ng of lz51-MuLV inlerspec I antigen (10,000 cpm per ng) and 5 ng of 1251-MuLV inlerspec II antigen (11,000 cpm per ng) as antigens and rabbit anti-FeLV serum, 1:3000 in Experiment A and I:600 in Experiment H 2s final dilutions. With these concentrations of antiserum the total antigen precipitated was 10,000 cpm of interspec I and 27,000 cpm of inlerspec II, 40 to 50yc of that added. The competing antigens FeLV interspec I and MuLV interspec II were added as indicated.
l3ecause the molecular weights of the interspec II polypeptides were similar to that of bovine serum albumin, the possible presence of the interspec II antigens in fetal calf serurn was tested by the competition radioimmunoassay.
There was no competition of fet,al calf serum with interspec II for antibody bindillg, tested to a 3000.fold greater concentration of serum proteins. We couclude that the interspec II pol?peptJides are uot coilstit.uents of fetal calf serum.
Comparison of Interspec I and /I by Monospecific Anfisera-Antisera prepared against the two purified proteins reacted only with their specific interspecies type, whell analyzed either by gel diffusion (Fig. 5) or radioimmuuoassay (Fig. 6), thus l)roviding further evidence that these autigeus represent two distinct classes of interspecies determinants.
Cell Fractionation and Separation of Interspec II from DXr;2 Polymerase-Another virus protein reactive with antisera directed against viruses of different species is the RKA-dependent DNA polymerase (13,14). For this reasoll, several experiments were carried out to distinguish interspec II polypeptides from the virus polymerase, especially as the mass of the enzyme has been reported as 70,000 daltons (37). Studies of the fracBtionation of interspec II polypeptides and the polymerase proteiu showed that they could bc completely sel)arated from one another ( Table I). All of the autigenic activity of the interspec II but moue of the polymerase was present in a supernatant fraction after freezing and thawiug the virus, whereas almost all of the polymerase and none of the antigeu remained in the particulate fraction after sonication. 2 In other experiments (data not shown) it was fouud that interspec II and the polymerase were antigenically distinct as well. Anti-interspec II serum with a titer of 1:12,000 against 5 llg of interspec II antigen by radioimmunoassay did not inhibit the polymcrase a&v&y of 3.4 pg of Rauscher virus even wheu added in vast excess (a final dilution of I:40 in the enzyme reaction mixture).
Noreovel interspec II failed to absorb antipolymerase IgG. As much as 0.8 pg of interspec II added with 3.4 pg total of virus protein did not show competition for antipolymcrase antibody uuder coilditions of limiting antibody in the reaction mixture.
11-c have 2 The extent of release of the interspec II polypeptides by freezing and thawing has been variable with different batches of virus, the ionic environment, and virlls concentration.
In every case, the proteins have been completely solubilized by combined freezing and thawing and sonication. 6. Radioimmunoassay analysis of anti-interspec I and anti-interspec II sera. The assay was performed as described under "Methods" with A, 5 ng of 1251-MuLV inlerspec II antigen (10,000 cpm per ng), or B, 2.5 ng of 12%MuLV interspec I antigen (18,600 cpm per ng). Rabbit anti-interspec I or rabbit antiinlerspec II serum was added as indicated.
concluded that interspec II is not an antigenic determinant of the virus DNA polymerase.
Species Distribution of Interspec II-The distribution of interspec II in different animal cell viruses was tested by use of the competition radioimmunoassay with '*"I-intwspec II and anti-FeLV antiserum (Table II). Murine and feline viruses contained equivalent amounts of the competing protein.
In contrast, none of the primate viruses or RD-114, at protein concentrations up to loo-fold greater than that of the murine and purified as previously described (16), was concentrated by high speed centrifugation, suspended in 20 mM Tris, pH 7.6,lOO mM NaCl, and 1 mM EDTA; 30 mg of total protein was dialyzed against 2 liters of 5 mM Tris, pH 7.6, and 1 mM EDTA for 2 hours (Fraction 1). The virus was then frozen in Dry Ice and methanol and thawed at 25" in a water bath, and the suspension centrifuged a.t 108,000 X g for 2 hours. The supernatant was retained (Fraction 2 supernatant), and the pellet (Fraction 2 pellet) was suspended in 5 mM Tris, pH 9.2, 1 mM EDTA, 1 mM dithiothreitol, 1 M KCI, and 20% glycerol, sonicated, and centrifuged at 108,000 X g for 2 hours. The soluble fraction was retained (Fraction 3 supernatant).
6 Interspec II was measured by competition radioimmunoassay as described under "Methods" with 5 ng of %I-interspec II antigen (16,000 cpm per ng) as antigen and a 1:450 final dilution of rabbit anti-FeLV serum. Before assay, Triton X-100, 0.4yo final ooncentration, was added to each fraction, except the sonicated pellet which already contained the detergent, and the fractions were incubated for 10 min at 37". The buffer solution used for the serial dilution of the antigen also contained 0.4% Triton X-100. The concentration of interspec II in the virus fractions was calculated by comparing the competition of the fraction to that of pure inlerspec II antigen.
c Virus RNA-dependent DNA polymerase was measured as described under "Methods." tein found in AMV, herpes virus, nononcogenic viruses, or mycoplasma.
Analysis of Uninfected Cells-In studies of the genetic origin of the interspec II antigen, analysis was made of uninfected cells to investigate the possibility that this protein is constitutively expressed in the host. Because of the possibility that interspec II is a membrane component, this experiment was also carried out with host cells infected with vesicular stomatitis virus, another membrane maturing virus. Under assay conditions which could detect interspec II at a concentration 10P4 of the total cell protein, none was found in the uninfected cell or cells productively infected with vesicular stomatitus virus (Fig. 7). In control experiments, a competing protein was readily detected in cells productively infected with Rauscher MuLV. DISCUSSION It may be hypothesized that the interspecies antigens of the mammalian animal cell RNA tumor viruses are components of proteins that are involved in specific functions common to these viruses and that have been conserved during evolution of difand feline viruses, showed any competition.
Nor was the proferent species These structures could be expected to include active sites of the RNA-dependent DNA polymerase, the RNA packaging protein or proteins, and the structural proteins that determine the configuration of the particles. Only three such proteins with interspecies homology are 110~ known, interspec I (I, Z), the RSA-dependent DXA polymerase (13,14), and interspee II, but there could IT-e11 be others present as minor components of the virus or not expressed as strong antigens.
It appears reasonable to expect that the maximum number of such prot,eills ix governed by the size of the virus genomc. Although genetic cl-idence of proteins coded for by the virus is available only for the DNA polgmerase (38), the selective nature of the interspcc alltigclCc determinants and the absence of any evidence for thpae proteiiis in uninfected cells strongly suggests that they WC virus coded. At this time the apparent mass of probable virus-qecific proteins is about 400,000 daltons, comprising the DNA1 polpmerase and the proteins containing the type-specific, species-qecifict, and interspee determinants (l-10, 37). As the total mass of the viral RSX is about lo7 daltons, many more virus proteins could bc accommodated if the genomc is non- After incubation for 60 min at 32" to allow virus adsorption, the medium was removed, 20 ml of Ml!%-10 were added, and incubation continued at 32" for 8 hours.
The medium was then removed and the cell sheets washed with 5 ml of PBS. The cells were removed with a rltbber policeman and suspended in 40 ml of PBS.
They were then concentrated by low speed centrifugation and suspended in 5 ml of fresh PBS. Uninfected JLS-V9 cells were propagated, mock infected with PBS, and processed in the same manner. repetitive.
However, if the genome is polyploid and contains 3 to 4 repeating 35 S subunits of 3.0 to 3.5 x lo6 daltons, as is suggested by genetic studies,3 then these proteins already appear to comprise the apparent limit of what might be coded for by such a subunit.
At this time the interspee II polypeptides and monospecific antiserum are useful reagents for virus characterization and for analysis of viral gene expression.
The species distribution of interspec II shows a strong homology between the murine and feline sarcoma or leukemia viruses, as is also demonstrated by interspec I and the polymerase protein.
It can be expected that rat and hamster viruses also contain the interspec II determinant because of the known similarity of these viruses (10, 12, 32).
No evidence of antigenic homology between interspec II and proteins of RD-114 and primate viruses was obtained from competition radioimmunoassay analysis. We believe, homever, that this lack of competition is due to a low affinity of the antifeline virus antibodies for interspec II proteins of primate viruses, not to the absence of the antigens.
Preliminary studies show that high concentrations of anti-woolly monkey virus antibodies precipitate the purified murine interspee IL protein.
Attempts to purify the analogous protein from woolly monkey and gibbon ape C type viruses are in progress.
Acknourledgnzenfs-abbe are indebted to Dr. Jack Gruber and the Office of Program Resources and Logistics, Xational Cancer Institute, through whose courtesy we were provided with the Rauschcr MuLT' and others of the C type viruses used in this investigation.
Thanks are also due to Mr. H. Silberstein for the propagation of vesicular stomatitis virus-infected JLS-V9 cells. We gratefully acknowledge the aid and advice of Prof. R. Porter and Drs. A. Williams and H. Anderson, Oxford University, in the application of the immunological techniques used in these studies.