Proc. Natl. Acad. Sci. USA Vol. 74, No. 3, pp. 1219-1223, March 1977 Immunology

Feline oncornavirus-associated cell-membrane antigen (FOCMA): Distinction between FOCMA and the major virion glycoprotein* (feline leukemia /tumor antigen/virion proteins/immunosurveillance) J. R. STEPHENSONt, M. ESSEXt, S. HINOt, W. D. HARDY, JR. §, AND S. A. AARONSONt t Laboratory of RNA Tumor , National Cancer Institute, Bethesda, Maryland 20014; f Department of Microbiology, Harvard University School of Public Health, Boston, Massachusetts 02115; and § Memorial Sloan-Kettering Cancer Center, New York, N.Y. 10021 Communicated by George Klein, December 27, 1976

ABSTRACT The humoral antibody response of feline leu- can be readily identified and quantitated in tissues and sera. kemia virus (FeLV)exposed cats to the feline oncornavirus- The present studies were, thus, undertaken to analyze sera from associated tumor cell-membrane antigen (FOCMA) is directly correlated with immunosurveillance against tumor development cats exposed to FeLV with respect to their serologic reactivities under natural conditions. By means of membrane immunoflu- against both FOCMA and the FeLV structural proteins gp7O orescence and radioimmunoprecipitation, the antibody re- and p30. The present results demonstrate that FOCMA is in- sponse to FOCMA was found to be independent of the antibody dependent of FeLV gp7O. In addition, the evidence supports response to the major envelope and core proteins of FeLV, gp70 and extends recent studies (20) that FOCMA differs from the and p3. This was especially true for healthy viremic cats,wbere FeLV major structural protein, p30. antigenemia with circulating FeLV gp7O and p30 apparently binds any free antibody to these proteins, but high levels of MATERIALS AND METHODS FOCMA antibody are often concurrently present. Exhaustive in vitro absorption of highly immune nonviremic serum with Virus. The Gardner strain of feline sarcoma virus gp70 and p30 also failed to remove FOCMA antibody activity. [FeSV(FeLV)] was obtained as sucrose-gradient purified virus These results indicate that FOCMA is not one of these major from Electronucleonics Laboratories, Rockville, Md. FeLV structural proteins. Sera. Sera from FeLV-exposed cats were obtained from Exposure of cats to the feline leukemia (FeLV) and sarcoma several sources, including two leukemia cluster households (14) (FeSV) viruses usually results in expression of the feline on- and the Angell Memorial Animal Hospital, Boston, Mass. (21). cornavirus-associated cell-membrane antigen (FOCMA) (1-5). The sera were classified as viremic or nonviremic on the basis This occurs both following laboratory inoculation with con- of a fixed cell fluorescence antibody test for FeLV proteins centrated live virus preparations (6-10) and after horizontal associated with blood leukocytes and platelets (22, 23). Serial contact exposure to virus under laboratory (6, 9, 11, 12) and field serum samples were obtained from a healthy 4-month-old cat conditions (13-15). The levels of FOCMA antibody detected obtained from a "FeLV-free" commercial colony (24) following in sera of virus-exposed cats have been shown to be predictive its introduction to a "leukemia cluster" household. All serum of tumor occurrence and growth. High FOCMA antibody titers samples were coded and assayed for antibody and antigen ac- appear to be protective against tumor development. Thus, the tivity in a double-blind fashion. Goat antiserum against immune response to FOCMA appears to function as an im- Tween-80-disrupted Theilen strain of FeSV (FeLV), pig anti- munosurveillance mechanism under natural conditions to goat IgG, and goat anti-cat IgG were provided by R. Wilsnack, prevent leukemia development following exposure to FeLV Huntington Laboratories, through the courtesy of J. Gruber, (3, 15). Office of Resources and Logistics, National Cancer Institute. The nature of FOCMA is as yet unknown. This antigen is Purification of FeLV gp7O and p30. Approximately 5 mg found in fresh biopsies of tumors etiologically associated with of purified FeSV(FeLV) was disrupted by incubation at 370 FeLV and/or FeSV and on FeLV-producing feline lympho- for 30 min with buffer containing 0.05 M Tris-HCI, pH 8.9, 0.01 blastoid tumor cells in culture (1, 16). Since sera containing M EDTA, 1.0 M NaCI, and 0.5% Triton X-100, and then cen- FOCMA antibody have generally been obtained from animals trifuged at 30,000 rpm for 30 min in a Beckman T40 rotor. The exposed to FeLV, the possibility that FOCMA represents a supernatant was dialyzed against 0.01 M N,N-bis(2-hydroxy- FeLV-structural protein(s) has been difficult to exclude. The ethyl)-2-aminoethanesulfonic acid-NaOH at pH 6.5, 0.001 M envelope glycoprotein of type C viruses, designated gp7O, is EDTA, and 0.1% Triton X-100 (BET buffer) at 40 for 18 hr, and expressed on the surface of type-C-virus-producing cells, and loaded on a 1.0 X 5.0 cm phosphocellulose column (Whatman, antisera to gp7O have been shown to be cytotoxic under ap- P11) equilibrated with BET buffer. After washing with 50 ml propriate conditions to type-C-virus-releasing cells (17, 18). of the same buffer, proteins were eluted with a 0.0-1.0 M linear Moreover, there has been a recent report implicating FeLV KCI gradient in BET buffer (100 ml) at a flow rate of 10 ml/hr. gp7O as FOCMA (19). The elution patterns of gp7O (0.15-0.20 M KCI) and p30 The development of radioimmunoassays utilizing specific (0.3-0.4 M KCI) were determined by radioimmunoassays de- type C viral structural proteins has made it possible to quanti- signed to detect interspecies antigenic determinants of each tatively measure antibodies to different viral proteins. More- protein (25, 26). Viral proteins were radiolabeled with 125I by over, in competition immunoassay, the respective viral proteins the method of Greenwood et al. (27). FOCMA Antibody Detection. The procedure for titration Abbreviations: FOCMA, feline oncornavirus-associated cell-membrane of FOCMA antibody by indirect membrane immunofluores- antigen; FeLV, ; FeSV, feline sarcoma virus; cence has been described in detail (6, 8). R-MuLV, Rauscher ; gp7O, major virus envelope Double Antibody Radioimmunoprecipitation and Com- glycoprotein; p3O, major virus core protein. petition Assays for FeLV gp7O and p30. For radioimmu- * This is number VII in a series; the preceding paper is ref. 40. noprecipitation assays, serial 2-fold dilutions of antisera were 1219 Downloaded by guest on September 23, 2021 1220 Immunology: Stephenson et al. Proc. Natl. Acad. Sci. USA 74 (1977)

>2000r . 2000-

1800 1Z 1600- E 0E

z I C4 1400[

I-8 0 z 0HI 1200- o *so 0 LU II -J 0 0 LU 6 10001 -J L 10 2 3. 40 5 I.to0 800s * moe 2~~ ~FATO NUBE 0 0 >'U J 0 600 to I~ ~ I CD * e * e e CU 400- z o 2001-

FRACTION NUMBER 00100o00o FIG. 1. Sodium dodecyl sulfate/polyacrylamide gel electropho- retic analysis of '251-labeled FeLV structural proteins. Around 30,000 . I. I. ,1 I cpm of 125I-labeled gp70 (0) and p30(0) were applied to individual 0 0 100 300 500 700 a 900oA 00 gels and electrophoresis was performed as previously described (39). Nanograms of FeLV 'jp7O per ml of plasma Molecular weight standards used for calibration, including bovine serum albumin (69,000), alkaline phosphatase (40,000), carbonic FIG. 2. Correlation between gp7O and p30 concentrations in sera anhydrase (29,000), fl-lactoglobulin (18,500), and lysozyme (14,300), of viremic (0) and nonviremic (0) cats. Each point represents one are indicated. individual cat. All cats were from FeLV exposure cluster environments (13-15). Viral antigen concentrations were determined by competition immunoassay as indicated in Materials and Methods and results are incubated with approximately i0,000 cpm of either 12I-labeled expressed as mean values from two separate determinations. FeLV gp7O or p30 for 3 hr at 370 and a further 18 hr at 40. For the assay of antibody to gp7O, the reaction mixture contained leukemia virus (R-MuLV) specifically precipitated the corre- in a volume of 0.2 ml: 0.01 M Tris-HCl at pH 7.8, 0.015 M sponding FeLV proteins, indicating that these proteins shared EDTA, 0.3% Triton X-100, 0.1% NaN3, 1% bovine serum al- antigenic determinants with known type C viral gp7O and p30. bumin, 0.001% normal cat serum, and 0.4 M NaCl. The con- As confirmation of the specificity of this binding, purified R- centration of NaCl was reduced to 0.01 M for assay of antibody MuLV gp7O, but not p30, competed for binding limiting anti- to p30. Goat antibody to cat IgG was added to a 1:10 dilution R-MuLV gp7O by the higher molecular weight 125I-labeled in a volume of 0.8 ml containing 0.01 M Tris-HCI at pH 7.8, FeLV protein. Conversely, R-MuLV p30 but not gp7O com- 0.001 M EDTA, 0.1% Triton X-100, 0.1 M NaCl, and incubation peted with anti-R-MuLV p30 for binding the 30,000 molecular was continued for 3 hr at 4°. The initial experiments established weight FeLV protein. The virus-specific nature of the two ra- that at this concentration of second antibody over 95% of cat dioiodinated FeLV proteins was demonstrated by the fact that IgG was precipitated even at the lowest dilutions (1:40) tested. FeLV, grown in cells of two different species, human and cat, After centrifugation at 2500 rpm for 15 min, the supernate was competed efficiently in- homologous immunoassays utilizing aspirated and the remaining radioactivity associated with the either of the '25I-labeled FeLV proteins. Thus, in the present immune precipitate was measured in a Searle 1285 gamma study, the FeLV proteins have been designated FeLV gp7O and counter. For competition radioimmunoassay, serial 2-fold p30, respectively, on the basis of their biochemical and im- dilutions of unlabeled antigen were incubated first for 1 hr at munologic similarities to the corresponding proteins of other 370 with a limited dilution of antibody to FeLV gp7O or p30 known mammalian type C RNA viruses. prepared in goats. The amount of unabsorbed antibody was In initial studies, sera from cats exposed to FeLV were tested then determined by procedures described above. for FeLV proteins by competition radioimmunoassay. As shown in Fig. 2, several of the sera tested lacked any detectable FeLV RESULTS gp7O or p30 as determined by radioimmunoassay. In other sera, Quantitation of FeLV gp7O and p30 in sera of FeLV- both viral proteins were present at readily detectable, and in exposed cats general coordinate, levels. These data correlate very well with previous studies by Hardy et al. in that sera expressing de- Radioiodinated FeLV viral proteins gp70 and p30 possessed tectable FeLV viral proteins by radioimmunoassay were also molecular weights of around 75,000 to 80,000 and 30,000, re- scored as viremic by the indirect fluorescence antibody test (22). spectively, by sodium dodecyl sulfate/polyacrylamide gel Similarly, cat sera that lacked detectable levels of either FeLV electrophoresis and exhibited over 95% radiochemical purity protein were previously designated as nonviremic. (Fig. 1). These were shown to be the major glycoprotein and structural proteins of FeLV by several criteria. Their molecular Immune response of FeLV-exposed cats -to FeLV weights were analogous to those of glycoproteins and structural structural proteins components of other mammalian type C viruses. Furthermore, It was of interest to analyze the same sera for presence of pre- antisera prepared against gp7O and p30 of Rauscher murine cipitating antibodies directed against the two radioiodinated Downloaded by guest on September 23, 2021 Immunology: Stephenson et al. Proc. Natl. Acad. Sci. USA 74 (1977) 1221

0 <32 000 0O0O 00 0 36400 0

6400 0 0 000 0 32 0 00 0 0

__ 3200[ 0o 00o 0 0 16 0 000 000 a) ._4 - * 0 0 1600 SO 0 00 > 8 oa a) 0

.0 .- 4 0 800 0 0 to ._ c 2(0 40* 00 2 0 0 U 200 00 -J (L) U. 100 Jo6W 8900 2 00** H *-1 *tMO540 100 200 400 800 1600 3200 0OO I: FeLV gp7O antibody titer g.0 FIG. 4. Correlation between antibody to FeLV gp7O and antibody to FOCMA in viremic (0) and nonviremic (0) cats. Titers are ex- 540 100 200 400 800 1600 3200 3200 pressed as reciprocals of the highest 2-fold dilution of serum giving FeLV gp7O antibody titer a positive result. FOCMA antibody was determined by membrane FIG. 3. Correlation between antibody to FeLV gp7O and p30 in immunofluorescence and FeLV gp7O antibody was determined by viremic (@) and nonviremic (0) cats, as determined by radioimmu- radioimmunoprecipitation. noprecipitation. Titers are expressed as reciprocals of the highest 2-fold dilution of serum at which 10% precipitation of the appropriate '251-labeled viral proteins occurred. Each point represents a mean for present at readily detectable levels in the majority of sera test- two separate determinations. ed. Kinetics of the immune response to FOCMA and FeLV FeLV viral proteins. As shown in Fig. 3, most of the sera that structural proteins in an uninfected cat contact- lacked detectable levels of FeLV structural antigens as deter- exposed to FeLV-infected cats contained antibody mined by competition radioimmunoassay As a further test of the above indication that antibody to capable of precipitating one or both FeLV proteins. The titers in from less than FOCMA differed from antibody to either FeLV gp7O or FeLV for antibody to FeLV gp7O this group ranged the kinetics of the immune response elicited against these to as as 1:3200. Antibody titers against FeLV p30 were p30, 1:40 high were For this a FeLV-nega- as as In each two antigens compared. purpose, somewhat higher, reaching levels high 1:12,800. containing FeLV- sera from these cats tive cat was introduced into a household case, when tested at low dilution, precipi- infected cats, and sera were obtained at intervals over the course over of the '251-labeled viral protein, tated 90% appropriate of the next several months. As shown in Table 1, antibody to ruling out the possibility that the cat sera were only recognizing in the labeled viral preparations. FOCMA was first detected 15 weeks prior to the development minor contaminants antigen of antibody titers to either gp7O or p30. These results In contrast, all sera from cats scored as viremic on the basis of significant the immune response to FOCMA is in- the detection of FeLV proteins in their sera, demonstrated no further indicate that these two FeLV structural measurable antibody to either FeLV gp7O or p30 (Fig. 3). These dependent of that directed against findings indicate that FeLV-exposed cats comprise two groups proteins. with respect to expression of virus and antibody to virus. Those Absorption of FOCMA-antibody-positive cat serum containing serum-associated FeLV antigens lacked evidence with FeLV gp7O or p30 or whose sera of antibody to either gp7O p30; conversely, those The availability of sensitive radioimmunoprecipitation assays no de- demonstrated viral antigens usually possessed readily made it possible to perform a final test to exclude a relationship or both. tectable antibody to either FeLV gp7O and p30 between antibody to FOCMA and antibody to FeLV gp7O or Lack of correlation between antibody to FOCMA and p30. Cat serum containing high-titered antibody to all three antibody to FeLV gp7O antigens was exhaustively absorbed with either purified gp70 or As in 2, under conditions where If antibody to FOCMA were the same as or closely related to p30. summarized Table directed it should be possible to correlate immunologic reactivity against gp7O was markedly reduced antibody against gp7O, against the presence of the two in individual serum samples. For these by absorption with FeLV gp70, the antibody titer both viremic and nonviremic cats were examined. As FOCMA was not significantly altered. Similarly, absorption studies, to shown in Fig. 4, there was some correlation between the pres- with FeLV p30 caused no detectable loss of antibody ence of antibody to FOCMA and antibody to FeLV gp70 in sera FOCMA. of nonviremic cats. This correlation, however, was not a very close one, because there were many sera in this group that DISCUSSION contained antibody to FOCMA in the absence of detectable The present study demonstrates that antibody to the FOCMA antibody to gp7O. The discrepancy between anti-gp7O and antigen is independent and distinct from antibody to the major anti-FOCMA titers was more striking in tests of sera from FeLV virion envelope glycoprotein, gp70, and the major virion viremic cats. Here, despite the absence of any detectable core protein, p30. Although we (16) and others (18) have de- antibody to FeLV gp7O (or to p30), FOCMA antibody was scribed the presence' o both membrane-associated p30 and Downloaded by guest on September 23, 2021 1222 Immunology: Stephenson et al. Proc. Natl. Acad. Sci. USA 74 (1977)

Table 1. Titers of antibody to FOCMA, FeLV gp7O, and p30 in a previously uninfected cat that was contact-exposed to FeLV-infected cats Weeks after exposure Test antigen 0 8 13 19 27 34 43 52 FeLV gp7O* <40 <40 <40 <40 <40 1280 640 1280 FeLV p3O* <40 <40 <40 <40 <40 640 320 640 FOCMAt 0 0 0 4 32 64 128 64 Data are given as the reciprocal of the highest 2-fold serum dilution yielding a positive result. * As determined by radioimmunoprecipitation. t As determined by indirect membrane immunofluorescence.

gp70 on FOCMA-positive feline tumor cells that produce brane immunofluorescence procedure. As shown in the present FeLV, several lines of evidence indicate a clear separation of studies, sera from nonviremic cats often do contain high titered these antigenic reactivities. Firstly, antibody to gp7O and p30 antibody to viral structural proteins as well as antibody to was never found in serum samples from viremic cats, even FOCMA. However, healthy viremic cats, in which viral pro- though such cats frequently exhibited high FOCMA antibody teins are circulating in vivo, lack detectable antibody to gp7O titers. Second, studies with sequential serum samples from an and often still contain high titered anti-FOCMA activity. FeLV-exposed cat illustrate that FOCMA antibody can arise Healthy viremic cats represent a significant fraction (up to 40%) independently from antibodies to gp7O and p30 in nonviremic of cats in leukemia cluster household populations, but are very animals. Third, viremic cats with high FOCMA antibody titers uncommon in the absence of constant FeLV exposure (14, 22). demonstrated levels of circulating viral antigens similar to those Whether the results of Ruscetti et al. (19) can be explained by in viremic cats that lacked detectable FOCMA antibody. This their not having tested such sera or whether their test for lack of correlation between amounts of virus antigens and "FOCMA" actually measures gp7O is not resolved. FOCMA antibody titer suggests that gp7O and p30 do not ab- The nature of FOCMA remains to be determined. It is pos- sorb FOCMA antibody in vivo. Finally, exhaustive absorption sible that this antigen may represent: (a) one or more low- of antiserum that had antibody to FOCMA, gp7O, and p30 with molecular-weight FeLV proteins expressed on membranes of purified gp7O or p30 markedly reduced antibody activity to infected cells; (b) a virion precursor polyprotein analogous to these virion structural proteins without reduction in FOCMA those recently described in cells infected with avian or murine antibody activity. oncornaviruses (33, 34); (c) a tumor-specific surface antigen The present findings are consistent with previous studies in coded for by the viral genome; or (d) a cell-coded protein that which FeLV-neutralizing antibody was shown to be distinct is specifically derepressed by FeLV. from FOCMA antibody (3, 4, 7, 10, 23, 28). In those studies, Certain investigators have suggested that cell surface antigens healthy viremic cats, which lacked detectable virus neutralizing on tumor cells infected with Friend murine oncornavirus rep- antibody, often exhibited high FOCMA antibody titers (3, 4, resent virion p15 (35, 36). In another study it was found that the 10). Available evidence indicates that virus neutralization can Moloney oncornavirus cell surface antigen (MCSA) was distinct be elicited by antisera prepared against purified viral gp7O from antigens that could be detected using goat or rabbit an- (29-31). However, this viral protein contains a broad range of tisera to Friend virus p15 or gp7O or Rauscher virus gp7O, p30, antigenic determinants detectable by immunoprecipitation, or p12 (37). The possibility that FOCMA is FeLV p15 is con- only some of which are involved in neutralization (26,32). Thus, sidered unlikely in view of the fact that only very minimal unlike the present studies that involved precipitation of 125I- amounts of this protein are present on reference target cells that labeled gp7O, the previous studies could not exclude the possi- are strongly positive for FOCMA (16, 17). Finally, as shown bility that FOCMA consisted of antigenic determinants of gp70 here, sera from healthy viremic cats often contained high-ti- distinct from those involved in neutralization. tered FOCMA antibody as well as large quantities of viral Recently, Ruscetti et al. reported that FOCMA antibody is structural proteins gp7O and p30. Such sera possessed no anti- indistinguishable from antibody directed against FeLV gp7O body to either of these proteins, presumably because of its ab- (19). These studies utilized an 125I-antiglobulin-cell binding sorption by the large viral antigen excess. It is probable that assay for titration of FOCMA rather than the standard mem- antibody to other viral structural components would also be absorbed under these conditions. Table 2. Presence of antibody to FOCMA FOCMA is present in large amounts on leukemia cells from in an immune cat before and after exhaustive absorption cats with induced or spontaneous tumors caused by FeLV, on with FeLV structural proteins gp7O and p30 fibrosarcoma cells from animals with virus-induced tumors, and on cells transformed in vitro by FeSV (refs. 1 and 16; Essex et Antibody titer* to: al., unpublished data). The antigen is not present on normal lymphoid cells or fibroblasts, even when these cells are infected Treatment of serum gp7ot p3Ot FOCMAt with FeLV but not transformed. The antigen is also not de- tectable on feline fibroblast cells transformed by murine sar- None 640 12,800 64 coma virus, regardless of whether or not such cells are super- Absorption with gp7O <40 6,400 64 infected with FeLV (Essex et al., unpublished data). Absorption with p30 640 40 64 Rohrschneider et al. (38) have recently isolated a tumor-specific *Reciprocal of highest 2-fold dilution giving a positive result. surface antigen from cells transformed by . tDetermined by radioimmunoprecipitation. This antigen is also distinct from the major viral proteins and, $Determined by indirect membrane immunofluorescence. thus, could be analogous to FOCMA. Downloaded by guest on September 23, 2021 Immunology: Stephenson et al. Proc. Natl. Acad. Sci. USA 74 (1977) 1223

Antibody to FOCMA is lytic for feline lymphoma cells in the 17. Hunsmann, G., Claviez, M., Moennig, V., Schwarz, H. & SchIfer, presence of complement, and levels of this antibody are directly W. (1976) Virology 69, 157-168. correlated with tumor immunosurveillance under natural 18. Schwarz, H., Hunsmann, G., Moenning, V. & Shafer, W. (1976) conditions (15). Thus, whether FOCMA is virus-coded, or a Virology 69, 169-178. virus-associated tumor-specific cellular antigen, this antigen 19. Ruscetti, S. K., Scolnick, E. M. & Parks, W. P. (1976) Proceedings an of the Cold Spring Harbor Meeting on Oncornaviruses (Cold provides important approach toward prevention of naturally Spring Harbor Laboratories, Cold Spring Harbor, NY), p. 45. occurring oncornavirus-induced cancers. 20. Charman, H. P., Kim, N., Gilden, R. V., Hardy, W. D., Jr. & Essex, M. (1976) J. Natl. Cancer Inst. 56, 859-861. This work was aided by Grants CA-13885, CA-18216, CA-16519, 21. Essex, M., Cotter, S. M., Hardy, W. D., Jr., Hess, P., Jarrett, O., and CA-15579 from the National Cancer Institute, Grant DT-32 from Mackey, L., Laird, H., Perryman, L., Olsen, R. G. & Yohn, D. the American Cancer Society, and Contract NCI-E-73-3212 from the S. (1975) J. Natl. Cancer Inst. 55, 463-467. National Cancer Institute. M.E. and W.D.H., Jr. are Scholars of the 22. Hardy, W. D., Jr., Old, L. J., Hess, P. W., Essex, M. & Cotter, S. Leukemia Society of America. We thank E. Jagher, M. Mandel, and (1973) Nature 244,266-269. E. Zuckerman for excellent technical assistance. 23. Hardy, W. D., Jr., Hess, P. W., MacEwen, E. F., McClelland, A. J., Zuckerman, E. E., Essex, M. & Cotter, S. M. (1976) Cancer Res. 1. Essex, M. (1975) Adv. Cancer Res. 21, 175-248. 36,582-588. 2. Essex, M. (1975) Pathobiol. Annu. 5, 169-196. 24. Essex, M., Cotter, S. M., Carpenter, J. L., Hardy, W. D., Jr., Hess, 3. Essex, M. (1976) Contemp. Top. Immunobiol. 6,71-106. P., Jarrett, W. & Yohn, D. S. (1975) J. Natl. Cancer Inst. 54, 4. Essex, M., Sliski, A., Hardy, W. D., Jr. & Cotter, S. M. (1976) 631-635. Cancer Res. 36, 640-645. 25. Stephenson, J. R. & Aaronson, S. A. (1973) J. Virol. 12, 564- 5. Essex, M. (1974) in Viruses, Evolution, and Cancer, eds. Kurstak, 569. E. & Maramorosch, K. (Academic Press, New York), pp. 513- 26. Hino, S., Stephenson, J. R. & Aaronson, S. A. (1976) J. Virol. 18, 541. 933-941. 6. Essex, M. & Snyder, S. P. (1973) J. Natl. Cancer Inst. 51, 27. Greenwood, F. C., Hunter, W. M. & Glover, J. S. (1963) Biochem. 1007-1012. J. 89, 114-123. 7. Jarrett, W. F. H., Jarrett, O., Mackey, L., Laird, H., Hardy, W. 28. Hardy, W. D., Jr. (1974) Vet. Clin. N. Am. 4, 133-146. D., Jr. & Essex, M. (1973) J. Natl. Cancer Inst. 51, 833-841. 29. Moennig, V., Frank, H., Hunsmann, G., Schneider, I. & Shafer, 8. Essex, M., Klein, G., Snyder, S. P. & Harrold, J. B. (1971) Nature W. (1974) Virology 61, 100-108. 233, 195-196. 30. Steeves, R. S., Strand, M. & August, J. T. (1974) J. Virol. 14, 9. Hoover, E. A., Olsen, R. G., Hardy, W. D., Jr., Schaller, J. P. & 187-189. Mathes, L. E. (1976) J. Natl. Cancer Inst. 56, in press. 31. Fischinger, P. J., Schafer, W. & Bolognesi, D. P. (1976) Virology 10. Schaller, J. P., Essex, M., Yohn, D. S. & Olsen, R. G. (1975) J. Natl. 71, 169-184. Cancer Inst. 55, 1373-1378. 32. Hino, S., Stephenson, J. R. & Aaronson, S. A. (1975) J. Immunol. 11. Essex, M., Klein, G., Snyder, S. P. & Harrold, J. B. (1971) Int. J. 115,922-927. Cancer 8, 384-390. 33. Stephenson, J. R., Tronick, S. R. & Aaronson, S. A. (1975) Cell 6, 12. Essex, M., Hardy, W. D., Jr., Cotter, S. M. & Jakowski, R. M. 543-548. (1975) in Comparative Leukemia Research, 1973, eds. Ito, Y. & 34. Naso, R. B., Arcement, L. J. & Arlinghaus, R. B. (1975) Cell 4, Dutcher, R. M. (Karger, Basel), pp. 431-436. 31-36. 13. Cotter, S. M., Essex, M. & Hardy, W. D., Jr. (1974) Cancer Res. 35. Friedman, M., Lilly, F. & Nathenson, S. G. (1974) J. Virol. 14, 34, 1061-1069. 1126-1131. 14. Essex, M., Jakowski, R. M., Hardy, W. D., Jr., Cotter, S. M., Hess, 36. Strand, M., Wilsnack, R. & August, J. T. (1974) J. Virol. 14, P. & Sliski, A. (1975) J. Natl. Cancer Inst. 54, 637-641. 1575-1583. 15. Essex, M., Sliski, A., Cotter, S. M., Jakowski, R. M. & Hardy, W. 37. Fenyo, E. M. & Klein, G. (1976) Nature 260,355-356. D., Jr. (1975) Science 190, 790-792. 38. Rohrschneider, L., Kurth, R. & Bauer, H. (1975) Virology 66, 16. Essex, M., de Noronha, F., Oroszlan, S. & Hardy, W. D., Jr. (1976) 481-491. in Proceedings of the Third International Symposium on De- 39. Laemmli, U. K. (1970) Nature 227,680-685. tection and Prevention of Cancer, ed. Nieburgs, H. E. (Marcel 40. Mathes, L. E., Yohn, D. S., Hoover, E. A., Essex, M., Schaller, J. Dekker, Inc., New York), in press. P. & Olsen, R. G. (1976) J. Natl. Cancer Inst. 56, 1197-1200. Downloaded by guest on September 23, 2021