Previous studies on Epstein-Barr virus (EBV)-positive B-cell lines have identified two distinct forms of virus latency. Lymphoblastoid cell lines generated by virus-induced transformation of normal B cells in vitro, express the full spectrum of six EBNAs and three latent membrane proteins (LMP1, LMP2A, and LMP2B); furthermore, these lines often contain a small fraction of cells spontaneously entering the lytic cycle. In contrast, Burkitt's lymphoma-derived cell lines retaining the tumor biopsy cell phenotype express only one of the latent proteins, the nuclear antigen EBNA1; such cells do not enter the lytic cycle spontaneously but may be induced to do so by treatment with such agents as tetradecanoyl phorbol acetate and anti-immunoglobulin. The present study set out to determine whether activation of full virus latent-gene expression was a necessary accompaniment to induction of the lytic cycle in Burkitt's lymphoma lines. Detailed analysis of Burkitt's lymphoma lines responding to anti-immunoglobulin treatment revealed three response pathways of EBV gene activation from EBNAl-positive latency. A first, rapid response pathway involves direct entry of cells into the lytic cycle without broadening of the pattern of latent gene expression; thereafter, the three "latent" LMPs are expressed as early lytic cycle antigens. A second, delayed response pathway in another cell subpopulation involves the activation of full latent gene expression and conversion to a lymphoblastoidlike cell phenotype. A third response pathway in yet another subpopulation involves the selective activation of LMPs, with no induction of the lytic cycle and with EBNA expression still restricted to EBNA1; this type of latent infection in B lymphocytes has hitherto not been described. Interestingly, the EBNA1+ LMP+ cells displayed some but not all of the phenotypic changes normally induced by LMP1 expression in a B-cell environment. These studies highlight the existence of four different types of EBV infection in B cells, including three distinct forms of latency, which we now term latency I, latency II, and latency III.

Exposure to the tiglian 12-O-tetradecanoylphorbol-13-acetate (TPA) represents one of the most efficient and widely used protocols for inducing Epstein-Barr virus (EBV)-infected cells from latent into lytic cycle. Since TPA is both a potent tumor promoter and a potent activator of the cellular protein kinase C (PKC), we sought to determine whether either of these activities was closely linked to EBV lytic cycle induction. A panel of TPA structural analogs, encompassing tiglians with different spectra of biological activities, was assayed on a number of EBV-positive B-lymphoid cell lines. Lytic cycle induction correlated with the capacity to activate PKC, not with tumor promoter status; some nonpromoting tiglians were as efficient as TPA in inducing lytic cycle antigen expression. We then sought more direct evidence for an involvement of PKC in the induction process. In initial experiments, 1-(5-isoquinolinyl sulphonyl)-2-methylpiperazine (H-7), the best available pharmacological inhibitor of PKC, completely blocked the induction of the lytic cycle by TPA and its active analogs. This is consistent with, but does not prove, a requirement for active PKC in the induction process, since H-7 targets PKC preferentially but also has some effects on other kinases. We therefore turned to the synthetic pseudosubstrate peptide PKC(19-36) as a means of specific PKC inhibition and to the closely related but inactive peptide PKC(19-Ser-25-36) as a control. Using the technique of scrape loading to deliver the peptides into cells of an adherent EBV-positive target line, we found that the pseudosubstrate peptide PKC(19-36) completely and specifically blocked tiglian-induced entry of the cells into the lytic cycle. The evidence both from TPA analogs and from enzyme inhibition studies therefore indicates that the pathway linking TPA treatment to lytic cycle induction involves active PKC. Interestingly, inhibition of PKC had no effect upon the spontaneous entry into lytic cycle which occurs in naturally productive cell lines, suggesting that spontaneous entry is signalled by another route.

Expression of the murine leukaemia virus (MLV) major Gag antigen p65(Gag) using the baculovirus expression system leads to efficient assembly and release of virus-like particles (VLP) representative of immature MLV. Expression of p180(Gag-Pol), facilitated normally in mammalian cells by readthrough of the p65(Gag) termination codon, also occurs efficiently in insect cells to provide a source of the MLV protease and a pattern of p65(Gag) processing similar to that observed in mammalian cells. VLP release from p180(Gag-Pol)-expressing cells however remains essentially immature with disproportionate levels of the uncleaved p65(Gag) precursor when compared to the intracellular Gag profile. Changing the p65(Gag) termination codon altered the level of p65(Gag) and p180(Gag-Pol) within expressing cells but did not alter the pattern of released VLP, which remained immature. Coexpression of p65(Gag) with a fixed readthrough p180(Gag-Pol) also led to only immature VLP release despite high intracellular protease levels. Our data suggest a mechanism that preferentially selects uncleaved p65(Gag) for the assembly of MLV in this heterologous expression system and implies that, in addition to their relative levels, active sorting of the correct p65(Gag) and p180(Gag-Pol) ratios may occur in producer cells.

No abstract

No abstract