More than 90% of adults are latently infected with Epstein-Barr virus (EBV), the causative agent of infectious mononucleosis, a self-limiting lymphoproliferative disease characterized by extensive T cell activation. Reactivation of this herpesvirus during immunosuppression is often associated with oncogenesis. These considerations led us to analyze the early events that occur after exposure of the immune system to EBV. Strong major histocompatibility complex (MHC) class II-dependent but not MHC-restricted, T cell proliferation was observed in vitro in response to autologous, lytically infected EBV-transformed B cells. By measuring the appearance of the early activation marker CD69 on individual T cell V beta subsets, we could demonstrate selective activation of human V beta 13- T cells. This was confirmed with murine T cell hybridomas expressing various human BV genes. While EBV- Burkitt's lymphoma cells were nonstimulatory, they induced V beta-restricted T cell activation after EBV infection. EBV specific activation was also demonstrated in cord blood cells, excluding a recall-antigen response. Thus, all of the characteristics of a superantigen-stimulated response are seen, indicating that induction of the EBV lytic cycle is associated with the expression of a superantigen in B cells. A model is presented proposing a role for the superantigen in infection, latency, and oncogenesis.

Short DNA sequences have been identified, originally in association with Kaposi's sarcoma (KS) biopsies, that are highly homologous to oncogenic, lymphotropic herpesviruses. Recently a virus, Kaposi sarcoma associated herpesvirus (KSHV) or human herpesvirus-8 (HHV-8), bearing these sequences has been identified in a cell line derived from a body cavity-based lymphoma. In this report, we show that the same sequences are present in KS biopsies as DNA molecules of a form and size characteristic of latent herpesviruses-large, covalently closed, circular episomes. The genomes migrate with an apparent size larger than the herpesvirus Epstein-Barr virus (172 kb). This form of the viral genome was found in four of four biopsies and three of five peripheral blood samples from KS patients. Linear forms of the viral genome, characteristic of viral replication, were not detected in the biopsies, but were present in the peripheral blood of three out of five patients. The sequences for KSHV/HHV-8 were also detected in the blood of four of five allograft patients and three of five healthy donors without KS suggesting that the virus is widespread throughout the human population.

Blast-1 is a human activation-associated glycoprotein expressed on the surface of leukocytes. Analysis of a translated sequence from a Blast-1 cDNA reveals a single hydrophobic sequence which could traverse the plasma membrane, but is devoid of charged residues that might represent a cytoplasmic tail. Consistent with this characteristic, Blast-1 is demonstrated here to be anchored to the cell surface through a glycosyl-phosphatidylinositol (GPI)-containing lipid. Comparison of Blast-1 to other GPI-anchored membrane proteins revealed a striking primary and secondary structure similarity with MRC OX45 and the lymphocyte function antigen LFA-3. The degree of overall amino acid sequence homology reveals that OX45 is a rat homologue of Blast-1. The greatest homology to LFA-3 occurs between their NH2-terminal Ig-like domains. Evidence is presented that demonstrates that Blast-1 and LFA-3 possess a disulfide-bonded second domain. These common characteristics demonstrate a structural and evolutionary relationship between Blast-1, OX45, LFA-3, and CD2, which in turn suggests a functional role for Blast-1 in cell-cell interactions in the immune response. The gene for Blast-1 has been localized to chromosome 1 q21-q23, indistinguishable from the CD1 cluster of Ig superfamily genes, raising the possibility that they may be linked.

Linear EBV genomes undergo a transition to the circular form characteristic of latency by 16-20 h post-infection. This transition requires that the infected cells be activated to the G1 stage of the cell cycle. Cellular proliferation and expression of the activation marker CD23 were not required. Nevertheless, 36 h post-infection, only cells expressing CD23 contained covalently closed, circular episomes (CCC), at an average of one copy per cell. Since the presence of CD23 at this time is predictive that a cell will immortalize, we suggest that the presence of CCC is required for CD23 expression and subsequent immortalization. The one CCC present in each CD23+ cell did not undergo amplification until well after the cells had acquired all of the characteristic phenotypic markers of immortalization. Therefore, while amplification is not necessary for proliferation and immortalization, circularization of a single genome is crucial to the establishment and maintenance of latency by EBV.

We have used Epstein-Barr virus (EBV) infection in vitro to delineate two distinct stages in B cell activation. Previous studies have shown that the BLAST-2 (EBVCS) (EBV cell surface) activation antigen is expressed on a small fraction of B cells within 24 h of stimulation with a variety of agents, including mitogens and EBV. In this study, we have been able to isolate the BLAST-2 (EBVCS)+ cells early after activation/infection with EBV. These cells are small B cells that are actively synthesizing RNA but not DNA, and are, therefore, clearly distinct from large proliferating lymphoblasts. In addition, they contain multiple copies of the EBV genome, express the viral nuclear antigen (EBNA) and, most importantly, proceed to undergo transformation when placed back in culture. By comparison, the BLAST-2 (EBVCS)- population does not undergo transformation, even though a fraction of these cells are activated for RNA synthesis and express EBNA. Thus, using the EBV system, we have been able to show directly that an activated B cell first expresses the BLAST-2 (EBVCS) antigen concomitant with an increase in RNA synthesis, and then subsequently proceeds to differentiate into a proliferating lymphoblast.

Studies have been performed on in vitro infection by Epstein-Barr virus (EBV) of subpopulations of human lymphocytes. B cells of adult peripheral or fetal cord blood transform with equal efficiency, whether assayed by DNA synthesis induction or by outgrowth of transformed lymphocytes. In contrast, unfractionated adult lymphocytes transform much less efficiently than those from fetal cord. Reconstitution experiments of different cell preparations indicated that this difference was due to a suppression of B-cell proliferation by adult Ig-negative lymphocytes which fetal Ig-negative lymphocytes were unable to perform. Separation of Ig-negative lymphocytes into various subpopulations revealed that the suppression was performed by T cells. Macrophages and null cells play little or no role in suppression. The relevance of this phenomenon to infection and recovery from EBV infection during and after infectious mononucleosis is discussed.