Purification and Immunological Characterization of the Major Internal Protein of the RD-114 Virus

Transcription

1 Proc. ivat. A cad. Sci. USA Vol. 69, No. 5, pp , Mlay 1972 Purification and Immunological Characterization of the Major Internal Protein of the RD-114 Virus (feline leukemia virus/c-type oncogenic viruses/rhabdomyosarcoma) S. OROSZLAN, D. BOVA,-M\. H. MARTIN WHITE, R. TONI, C. FOREMAN, AND R. V. GILDEN Flow Laboratories, Ine., Rockville, Maryland Communicated by Robert J. Heubner, February 28, 1972 ABSTRACT The major internal protein of the RD-114 virus that appeared in a human tumor cell line, RD, after passage through fetal cats was purified by isoelectric focusing and used to prepare antibody in guinea pigs. The protein has a molecular weight of 33,500 in sodium dodecyl sulfate-polyacrylamide gels, and an isoelectric point of 9.1. This contrasts with a molecular weight of 25,000 and an isoelectric point of 8.3 obtained previously for the major protein of cat C-type viruses. The RD-114 protein carries determinants common to all mammalian C-type viruses but not species-specific determinants of cat, mouse, rat, hamster, chicken, and Russell's viper C-type viruses. In addition, two other proteins found in cat virus preparations were not detected in the RD-114 virus. Antiserum to the RD-114 viral protein proved highly specific by complement fixation and gel diffusion assays, and, in addition to the above mentioned mammalian viruses, the antiserum did not react with concentrates of the Woolly monkey or Gibbon ape C-type viruses. These data support the conclusion that the RD-114 virus is a human virus activated by passage through fetal cats. Until additional isolates are made from human cells, the identity of the RD-114 virus as a human C-type tumor virus cannot be fully established; however, this would be highly probable if the species-specific group-specific antigen of RD-114 could be demonstrated in human tumor and/or embryonic tissues. Abbreviations: FeLV, feline leukemia virus; M\IuLV, murine leukemia virus; gs, group specific; CF, complement fixing; SDS, sodium dodecyl sulfate Recently, AMcAllister et al. (1) isolated a C-type virus from a human tumer (rhabdomryosarcoma) cell line (RD) after passage through fetal cats. The virus (RD-114) was differentiated from feline C-type viruses (FeLV) on antigenic grounds (2), based mainly on lack of FeLV-specific determinants associated with the major internal protein of all FeLV isolates (3). These determinants are designated as group specific (gs)-1 based on the nomenclature of Gregoriades and Old (4), first suggested for a purified component from mouse C-type viruses. Since a homologous protein based on molecular size, percent composition in the virion, and endgroup analysis (7) is found in all mammalian C-type viruses, we have retained the gs-1 designation to refer to speciesspecific determinants associated with this protein of the RD- 114 virus. The virus does contain crossreactive determinants (2), originally termed gs-3 (5), which are also located on the major internal protein (gs protein) of all mammalian C-type viruses studied to date (3, 6-8). The gs-3 determinants have been found in all mammalian C-type viruses including two recently described primate viruses (9-12). Since the speciesspecific determinants can be used to discriminate between various C-type viruses (3, 6, 7), one requirement for ultimate species classification of RD-114 is the preparation of the antiserum. This paper describes the purification of the major internal protein of RD-114 and further documents immunological differences between this virus and other C-type viruses. MATERIALS AND METHODS All virus preparations were obtained from chronically infected cultures. The RD-114 virus and cell line were obtained from Dr. R. iicallister and grown in this laboratory in 946-ml (32-oz) bottles as monolayer cultures. The Gibbon ape and Woolly monkey viruses (9-11) were derived from material originally provided by Dr. T. Kawakami (University of California, Davis) and transferred to human cell monolayers. We thank Dr. W. P. Parks (NCI) and Dr. Paul Price (Microbiological Associates, Inc.) for providing us with the latter cultures. Methods of virus harvest, purification, and concentration by continuous-flow isopycnic centrifugation, isoelectric focusing, polyacrylamide gel electrophoresis, gel diffusion, and complement fixation have been described in previous studies on other C-type viruses (3, 13-15). Radioimmunoassays were performed with purified gs protein of FeLV, iodinated by the chloramine-t method of Hunter (16), to give about 1 atom of iodine per molecule of gs protein and a specific activity of about 4.5 X 10-7 Ci (106 dpm)/ g of protein. The antisera used were prepared in this laboratory, with the exception of the goat antiserum to FeLV, which was obtained from Dr. Roger Wilsnack (Huntingdon Laboratories, Baltimore, \Id.). This antiserum was prepared against purified FeLV disrupted by Tween-ether. RESULTS Recently (7), we summarized the evidence for a distinct species-specific antigenic reactivity associated with the major internal protein of C-type viruses from six animal species, i.e., mouse, hamster, cat, rat, chicken, and Russell's viper. The three relatively recent isolates now under study, Woolly monkey, Gibbon ape, and RD-1 14 viruses, have not reacted with antisera that are specific for viruses from these six species. These viruses do contain the intermammalian crossreactive determinant, gs-3, as shown (2), as in Fig. 1 for RD-1 14, and as described elsewhere for the Woolly monkey and Gibbon ape viruses (11, 12, Kawakami and Gilden, unpub-

2 1212 Immunology: Oroszlan et al. FIG. 1. Gel-diffusion analyses of RD-114 virus with specific antisera to (upper set) viper (VSW), chicken (AMV), and mouse (MuLV); (lower set) cat (FeLV), hamster (HaLV), and rat (RaLV). In the center, the gs-3 reaction can be seen; two different antisera [rat antiserum to mouse sarcoma virus (MSV)] and goat antiserum to FeLV were used. Note that in both cases, only single joined lines are formed, while when tested against the homologous antigens, either multiple lines are seen (goat in Fig. 2) with distinct spurs or single lines with distinct spurs (6). This indicates that the gs-3-containing sera contain, without exception in our experience, gs-1 antibodies as well. This is as expected if both gs-1 and gs-3 determinants are on the same molecule and if the gs-1 determinants are the stronger immunogens. B, buffer control, throughout. lished). The reactions obtained with the goat antiserum to FeLV are especially significant since this serum detects three physically separable antigenic components in disrupted FeLV. As shown in Fig. 2, this serum reacts with only one component in RD-114 virus, and in this case with a TABLE 1. Failure of RD-114 antigen to inhibit precipitation of gs antigen of FeLV by antiserum to gs-1 antigen of FeLV CF units/ Precipitated 0.1 ml of gs [1251] % homologous Antigen Inhibi- Dilution of test virus system (cpm) tion FeLV 1: : : :10, MuLV 1: RD-114 1: No virus These tests were performed by the double-antibody method with a 1:1600 dilution of specific antiserum to FeLV gs-1 and about 10 ng of labeled gs antigen. Under these conditions, 50% of the label was precipitated in the control incubations (no virus added) by addition of goat antiserum to guinea pig gammaglobulin. Prior incubation of antiserum to gs-1 with Tween-ether disrupted unlabeled virus resulted in complete inhibition of precipitation only if FeLV was used. Neither MuLV nor RD-1 14 preparations gave detectable inhibition of precipitation. The sensitivity of the inhibition procedure is about 0.01 CF units/ 0.1 ml (30% inhibition). Proc. Nat. Acad. Sci. USA 69 (1972) reaction of partial identity when compared to FeLV; this partial reaction is, however, "identical" with the precipitin reactions formed with the other heterologous C-type viruses. This partial reaction defines the gs-3 determinants, while the excessive reaction with the same physical component clearly indicates the existence of species-specific (gs-1) determinants on this component. In other studies (data not shown), the goat antiserum produced a single precipitin line with highly purified gs protein of FeLV that was continuous with the line produced against this antigen by either guinea pig antiserum to FeLV gs-1 antigen or rat antiserum to MuLV containing antibodies to gs-3 antigen, when either of these sera were placed adjacent to the goat serum. This is consistent with and supports previous conclusions that gs-1 and gs-3 determinants are present on the same molecule (3, 6, 7). Further quantitative evidence for the lack of FeLV gs-1 determinants in RD-114 was provided by radioimmunoprecipitation-inhibition tests (17). A guinea pig antiserum to FeLV gs-1 antigen was diluted to give about 50% precipitation of about 10 ng of a purified preparation of FeLV gs protein labeled with iodine-125. The serum was then incubated at 370 with unlabeled, disrupted viruses before addition of the labeled antigen. Displacement of the label from precipitates to the free antigen position in sucrose gradients was taken as a measure of inhibition. 30% Inhibition was taken as an endpoint; the sensitivity of the inhibition tests was 0.01 CF unit/ 0.1 ml for FeLV. Based on a value of 3 CF units/,gg of protein per ml previously determined for purified gs proteins (13, 14), this test readily detects 3 ng of antigen. In several experiments, neither MuLV nor RD-114 preparations showed any inhibition, even when microgram amounts of antigen (based on homologous CF tests as well as protein determinations) were used (Table 1). The above results constitute strong negative evidence that RD-114 is not a feline or other of the well-studied C-type _.~~~~~. FIG. 2. Gel-diffusion analyses of FeLV, RD-1 14 virus, MuLV, and HaLV with a polyspecific goat antiserum to FeLV (1S-8). Viruses were disrupted by Tween 80-ether and placed in the outer wells. Protein concentrations were matched at about 0.3 mg/ml. The goat antiserum (center well) was prepared by Dr. Roger Wilsnack by multiple inoculations of ether-disrupted FeLV. The partial identity line between FeLV and both MuLV and RD-114 (spurs) represents the gs-3 reaction. The additional reactivity of the innermost component represents gs-1 determinants, as shown by tests with antisera to gs-1 antigen or purified gs antigen in appropriate patterns (data not shown). The two additional FeLVspecific components can also be detected by tests with individual guinea pig sera, and are physically separable from the innermost line.

3 Proc. Nat. Acad. Sci. USA 69 (1972) viruses. Positive evidence for this conclusion was provided by immunochemical characterization of the major internal antigen of the RD-114 virus. The virus was labeled with [3H]- aminoacids, disrupted by a method designed to give rapid solubilization of viral proteins (unpublished data), and subjected to isoelectric focusing. Fractions were tested for radioactivity and for gs-3 reactivity with two different antisera: rat antiserum prepared against mouse sarcoma virus (MSV) (6) and goat antiserum to FeLV. The significant feature of this separation (Fig. 3), now repeated several times with concordant results, was the coincidence of a clear radioactive zone with a peak at ph 9.1 with gs-3 reactivity in gel diffusion by both test antisera. Only two fractions showed reactivity and these were pooled and used to immunize guinea pigs. Animals were inoculated 3-4 times at weekly intervals with interspersed trial bleedings. Results obtained with one serum, obtained in large volume after positive findings, will be described; however, preliminary tests with sera from several other inoculated animals show comparable results. In gel diffusion assays (Fig. 4), the antiserum to RD-114 proved highly specific for RD-1 14; it showed a single precipitin line continuous with the purified antigen used for immunization. No reactions were obtained with any of the other C-type viruses, including the Gibbon ape and Woolly monkey viruses. Preparations of these viruses did contain gs-3 reactivity, providing some assurance that sufficient viral mass was present to detect a crossreaction with the antiserum to RD-1 14 gs antigen. In complement-fixation tests, with 4 units of antiserum, this pattern of specificity was retained (Table 2) and, C) x E Ca u 1- FRACTION NUMBER FIG. 3. Isoelectric focusing of RD-114 virus proteins in a ph gradient of The electrofocusing column (LKB 8101, 110 ml) contained 2% Ampholine and a urea gradient of 0-6 M. Purified [3H]aminoacid-labeled and unlabeled virus were mixed (total counts: 2 X 105 cpm; total protein: 2 mg), and a viral lysate was prepared for electrofocusing according to a newly developed method (to be published) based on disruption of virus with sodium dodecyl sulfate (SDS). The detergent was completely removed by passage through a Dowex 1-X2 ion-exchange column (18), to which all the nucleic acids were also bound. Electrofocusing was performed at 300 V and 60 for 48 hr. Fractions of about 2 ml were collected and, after removal of urea by dialysis, aliquots were assayed for radioactivity (13, 14) in a Beckman LS- 230 liquid scintillation system, and tested in immunodiffusion with MSV-I and a goat antiserum prepared against FeLV disrupted with Tween 80-ether. Both sera detected the shared mammalian C-type gs antigen (gs-3). Positive reactions were found only in fractions 35 and 36 (see inset), coinciding with a major radioactive peak at ph 9.1. ph FIG. 4. Major Internal Protein of the RD-114 Virus 1213 Gel-diffusion analyses of guinea pig antiserum to the major gs protein of RD-144 virus. Purified MuLY, FeLY, RaLY, and RD-i114 proteins, and disrupted RD-114 virus matched at a concentration of about 0.3 mg of protein/nil, were tested against guinea pig antiserum to RD-i 14 purified proteins. As shown, this serum reacted with a single component in disrupted RD-i 14 that was immunologically identical to the RD-114 purified protein (as expected). In addition to the results shown, concentrates (not purified proteins) of MuLV, FeLY, RaLY, HaLY, Woolly monkey, and Gibbon viruses did not react with this serum. in addition, the serum appeared highly specific for screening of infected and normal cell extracts. This provides a method TABLE 2. Specificity of guinea pig antiserum to RD-114 gs antigen CF titer versus CF titer antiserum versus to homologous Test antigen RD-114* antiserum RD-114 virus Purified gs antigen of RD virus HaLV <2 128 FeLV <2 256 MuLV <2 256 RaLV <2 128 Gibbon virust <2 N.A. Woolly monkey virust <2 N.A. Cell packs: (20% V:V) Infected human cells with RD-114 virus Infected hamster cells with HaLV <2 8 Infected mouse cells with MuLV <2 32 Infected cat cells with FeLV <2 64 Infected rat cells with RaLV <2 16 RD-human cells <2 <2 ESP-human cells (19) <2 8t * Used at 1: 40-1 :80 dilutions, 4-8 units based on homologous titration with purified RD gs antigen. t Both virus preparations contained gs-3 determinants when assayed in gel diffusion. t Tests with anti-mulv gs serum. N.A. = Not available. HaLV, hamster leukemia virus; RaLV, rat leukemia virus.

4 1214 Immunology: Oroszlan et al. "4*:. MuLV alp& FeLV 0 i 0 I I l RD114 A W epa FIG. 5. Polyacrylamide gel electrophoresis of MuLV, FeLV, and RD-114 virus, and their purified gs proteins. Disrupted virus and purified gs protein were analyzed in SDS-polyacrylamide gels [10% acrylamide-0.1% SDS-0.1 M sodium phosphate buffer (ph 7.0)]. Before electrophoresis, the virus and gs protein were treated with SDS (1%), urea (6 M), and 2-mercaptoethanol at 370 for 2 hr. The samples were then dialyzed against 0.1 M phosphate buffer (ph 7.0) containing 0.1% SDS-0.5 M urea-0.1% 2-mercaptoethanol. Electrophoresis (from top to bottom) was for 5 hr at 8 ma, per gel, and gels were stained with Coomassie blue. The purified gs proteins have calculated molecular weights of 31,000 (MuLV); 25,000 (FeLV); and 33,500 (1D-114). for measurement of virus growth in various cell lines based on development of antigenic mass. Our preliminary tests have shown no evidence of gs antigen of RD-1 14 in several human cell preparations, including the virus-free RD cells and the human ESP-1 cell line (19) that contains a virus with MuLV gs-1 determinants (17), and RNA-dependent DNA polymerase activity that is antigenically indistinguishable from MuLV polymerase (12). The purified protein from RD-1 14 was also analyzed by electrophoresis in SDS-polyacrylamide gels. Fig. 5 is a composite figure comparing electrophoretic patterns of MuLV, FeLV, and RD-114, and the major internal protein isolated from each virus by isoelectric focusing. As shown, the pattern of the disrupted viruses are quite complex when analyzed by staining methods. This contrasts with the relatively small number (about 5) of polypeptides detected by use of radioactive virus (3, 14). When the radioactive zones were analyzed in acrylamide gels, multiple bands were resolved from the acidic region (ph about 4.5) of the electrofocus separation, while only a single band was generally obtained from peak fractions in the neutral or alkaline regions of the electrofocus separations of the various viruses. Despite the differences in isoelectric points (MuLV, 6.7; FeLV, 8.3; RD-114, 9.1), the proteins are of very similar molecular weight (MuLV, 31,000; FeLV, 25,000; RD-114, 33,500), and are obviously the major proteins by percent composition of the various viruses. When high concentrations of the purified RD-114 gs protein were analyzed, as in Fig. 5, additional stained bands were seen in the higher molecular weight region of the gels. Whether these are aggregates or impurities is not clear; however, they were not present in all preparations. W N;W. Proc. Nat. Acad. Sci. USA 69 (1972) DISCUSSION The RD-1 14 virus possesses all the biophysical and biological attributes of known C-type viruses (2), including the ability to rescue the (MSV) genome from nonproducer (MSV-induced) hamster tumor cells (Long C., Lee Y. K. & Gilden, R. V., unpublished). The presence of species-specific, as well as crossreactive, determinants on the major internal virion protein is another interesting feature of the viruses. This suggests a functional constraint on the gs-3 region of the molecule during its evolution in mammals from a hypothetical common ancestral C-type virus or gs gene (20, 21), which we postulate must have existed at some time before the divergence of reptiles, birds, and mammals (22, 23), i.e., at least 300 million years ago. The significant observation reported here is that the RD-114 virus represents a new species distinguishable from viruses of six other mammalian species. The detailed distinction from cat and other known C-type viruses by reciprocal antisera increases the probability that the virus is of human origin. This conclusion is further strengthened by the observations (12) that the RD-1 14 RNA-dependent DNA polymerase was not inhibited by antisera that inhibit enzymes from FeLV, murine (MuLV), hamster (HaLV), and rat (RaLV) leukemia viruses. In similar tests, the Woolly monkey and Gibbon ape virus polymerases were not inhibited by the antisera (12); thus, RD-114 apparently falls into a "primate group" of C-type viruses. The need for an antiserum to RD- 114 polymerase for establishment of the reality of such a primate virus group, based on this property, is obvious. Availability of the specific antisera to the gs protein of RD-1 14 virus provides an opportunity to test for the presence of this protein in human materials. We thank Dr. Wilsnack for providing the goat antiserum (prepared under the auspices of the Resources and Logistics Section of the Special Virus Cancer Program, Contract NIH-69-54) and Dr. W. P. Parks (NCI) for calling our attention to its existence. We acknowledge with thanks the technical assistance of William Clagett, Michelle Kershes, Walden Saunders, Larry Masters, Annie Chien, and Steve Rosenthal. This work was supported by Contract NIH-NCI of the Special Virus Cancer Program of the National Cancer Institute, National Institutes of Health, Bethesda, Md. 1. McAllister, R. M., Nelson-Rees, W. A., Johnson, E. Y., Rongey, R. W. & Gardner, M. B. (1971) J. Nat. Cancer Inst. 47, McAllister, R. M., Nicolson, M., Gardner, M. B., Rongey, R. W., Rasheed, S., Sarma, P. S., Huebner, R. J., Hatanaka, M., Oroszlan, S., Gilden, R. V., Kabigting, A. & Vernon, L., (1972) Nature New Biol. 235, Oroszlan, S., Huebner, R. J. & Gilden, R. V. (1971) Proc. Nat. Acad. Sci. USA 68, Gregoriades, A. & Old, L. J. (1969) Virology 37, Geering, G., Aoki, T. & Old, L. J. (1970) Nature 226, Gilden, R. V., Oroszlan, S. & Huebner, R. J. (1971) Nature New Biol. 231, Gilden, R. V. & Oroszlan, S. (1972) Proc. Nat. Acad. Sci. USA 69, Schafer, W., Lange, J., Bolognesi, D. P., de Noronha, F., Post, J. E., & Rickard, C. G. (1971) Virology 44, Theilen, G. H., Gould, D., Fowler, M. E. & Dungworth, D. L. (1971) J. Nat. Cancer Inst. 47, Kawakami, T. G., McKain, D. & Fielding, H. (1972) Fifth International Symposium on Comparative Leukemia Research (Karger, Basle), in press. 11. Kawakami, T. G., Huff, S. D., Buckley, P. M., Dungworth, D. L., Snyder, S. P. & Gilden, R. V. (1972) Nature New Biol. 235,

5 Proc. Nat. Acad. Sci. USA 69 (1972) 12. Scolnick, E., Parks, W., Todaro, G. J. & Aaronson, S. (1972) Nature New Biol. 235, Oroszlan, S., Fisher, C. F., Stanley, T. B. & Gilden, R. V. (1970) J. Gen. Virol. 8, Oroszlan, S., Foreman, C., Kelloff, G. & Gilden, R. V. (1971) Virology 43, Gilden, R. V., Oroszlan, S. & Huebner, R. J. (1971) Virology 43, Hunter, R. (1969) Proc. Soc. Exp. Biol. Med. 133, Gilden, R. V., Parks, W. P., Huebner, R. J. & Todaro, G. J. (1971) Nature 233, Weber, K. & Kuter, D. (1971) J. Biol. Chem. 246, Major Internal Protein of the RD-114 Virus Priori, E. S., Dmochowski, L., Myers, B. & Wilbur, J. R. (1971) Nature New Biol. 232, Payne, L. N. & Chubb, R. C. (1968) J. Gen. Virol. 3, Taylor, B. A., Meier, H. & Myers, D. D. (1971) Proc. Nat. Acad. Sci. USA 68, Huebner, R. J. & Gilden, R. V. (1972) in RNA Viruses and Host Genomes in Oncogenesis (North Holland, Amsterdam), in press. 23. Gilden, R. V. & Huebner, R. J. (1972) in RNA Viruses and Host Genomes in Oncogenesis (North Holland, Amsterdam), in press.