The ZEBRA protein is a transcriptional activator that induces expression of viral lytic genes in cells harboring latent Epstein-Barr virus (EBV). In this report it is shown that a derivative of ZEBRA that cannot activate transcription (Zd) can be used to detect and characterize activation domains. Three expression vectors that allow the fusion of putative activation regions in any reading frame were constructed using Zd. These vectors were used to demonstrate the activity of different classes of activation domains using a chloramphenicol acetyltransferase (cat) reporter gene construct containing seven ZEBRA response elements (Z7). The Zd/Z7 system effectively detected proline-rich, glutamine-rich and acidic activation domains in a variety of cell lines and cell types. Using a bioassay unique to the EBV Zd/Z7 system, fusion constructs can also be tested for the ability to activate gene expression directly from a chromatin structure, the EBV genome. These studies indicate that the Zd/Z7 system is an alternative to GAL4 and can be a useful tool for identifying heterologous activation domains.

Latent infection of B lymphocytes by Epstein-Barr virus (EBV) can be disrupted by expression of the EBV ZEBRA protein. ZEBRA, a transcriptional activator, initiates the EBV lytic cascade by activating viral geneexpression. ZEBRA is also indispensable for viral replication and binds directly to the EBV lytic origin of replication. The studies described herein demonstrate that the activation domain of ZEBRA is not unique and can be replaced by a heterologous acidic, proline-rich, or glutamine-rich activation domain. ZEBRA activation domain swap constructs retain ZEBRA's native abilities to activate specific EBV promoters, to disrupt EBV latency, and to stimulate replication at the EBV lytic origin. Additional work, employing sequential and internal deletions of ZEBRA's N-terminal activation domain, indicates that its separate activities are not attributable to specific subdomains but are spread throughout its N terminus and therefore cannot be inactivated by deleting localized regions. MATERIALS AND METHODSCells. The B-cell lymphoma lines Clone 16 (CL16), Raji, BL41/CL16, and BJAB were maintained in RPMI medium supplemented with 8% fetal calf serum and antibiotics at 37°C under a 5% CO 2 atmosphere (4). CL16 cells are human B lymphocytes which harbor the nontransforming P3HR-1 strain of EBV in the latent state (36). Raji cells are human B lymphocytes which contain a deleted EBV episome that can be induced to express early antigen, but which cannot replicate (24). The BL41/CL16 cell line is a derivative of the EBV-negative Burkitt's lymphoma BL41 cell line, which has been stably infected with the EBV CL16 virus. After prolonged culture in vitro, the EBV in BL41/CL16 cells has become progressively more tightly latent and can no longer be induced by * Corresponding

No abstract