Brie£ De£initive Report Amplified Env and Gag Products on Akr Cells Origin from Different Murine Leukemia Virus Genomes* by Jwu-sheng Tung and Erwin Fleissner

Products of both the env (envelope) and gag (core) genes of murine leukemia viruses (MuLV) are present on the plasma membrane of AKR mouse thymocytes (1, 2). The env product is gp70. The gag products are the polyproteins gP85 and gP95. The expression of antigens associated with both gp70 and the polyproteins on AKR thymocytes becomes amplified before the onset of overt leukemia (3). In this report, we give evidence that the env information for the amplified gp70, and the gag information for the amplified gP85 and gP95, are derived from different proviral genomes. The serum anti-X. 1 (4) precipitates one of at least two varieties of gp70 present on AKR thymocytes (5). The type variant precipitable by anti-X. 1 serum was called Xgp70 (5), but we now propose to name it Ec ÷ gp70 because anti-X. 1 serum recognizes gp70 only from ecotropic virus (6), and also because peptide maps of X-gp70 of the plasma membrane resemble peptide maps of gp70 of ecotropic virus (6). Accordingly, the gp70 that remains after Ec + gp70 has been eliminated by precipitation with antiX. 1 serum is referred to below as Ecgp70. The peptide maps of Fig. 1 emphasize the contrast between AKR cellular Ec ÷ gp70 (maps b and e), which resembles ecotropic viral gp70 (6), and AKR cellular Ecgp70 (maps c and f), which resembles xenotropic viral gp70 (6). The peptide map for total gp70 (a and d) is clearly composed of Ec ÷ gp70 and Ecgp70. In qualitative terms, the maps for thymocytes of 2-mo-old AKR mice (Fig. 1 A) and those for AKR leukemia cells (Fig. 1 B) are rather similar. Fig. 2 illustrates that Ecgp70 is amplified with age but Ec ÷ gp70 is not, implying that the env gene involved in amplification does not belong to ecotropic virus. With regard to the polyproteins, the following evidence implies that the polyproteins gP85 and gP95 are products of one gag gene, and not of alternative gag genes: First, they have never been observed independently of one another. Secondly, as shown in Fig. 1 C, gP85 and gP95 have similar peptide maps (g and h) with the exception of a prominent gP95 spot (compare g and h) that is also found in the map of p l0 purified from virions (i); we do not find this spot in maps of the other processed gag proteins, p30, p15, and p12. These observations are in accord with our finding that no typespecific or group-specific anti-gag-protein serum other than anti-pl0, can distinguish gP85 from gP95. Because the p l0 sequence is encoded at the 3'-end of the gag gene (7), we conclude that both polyproteins are derived from the same viral gene, and