Kaposi's sarcoma (KS),Kaposi's sarcoma (KS)-associated herpesvirus (KSHV) (or human herpesvirus 8) is consistently associated with all epidemiologic forms of KS and is recognized as the etiologic agent of the disease (15,27). Within the KS lesion, KSHV infects the spindle-shaped cells that characterize the tumor as well as their endothelial cell precursors (15,62,100). The majority of infected cells harbor the KSHV genome in a latent form, with a small percentage entering a lytic cycle and producing infectious virus (82,102,116).KSHV genes with the potential to deregulate cellular growth have been described previously. Several of these genes have homology to human oncogenes and growth factors or transforming genes of other oncogenic herpesviruses, while others are unique to KSHV (33). KSHV regulatory gene products include homologs of cellular cytokines (116), which has been shown to have transforming ability (73); and the latent nuclear antigen (LANA/ORF73), which modulates cellular transcription (84). Interestingly, three of these gene products, LANA, vFLIP and vCYC are thought to constitute the latent gene expression program in KSHV and are consistently expressed in all virally infected cells in KS, primary effusion lymphoma (PEL), and multicentric Castleman's disease (MCD) (34,36,102). Other gene products such as the vGCR and vIL-6 are expressed in a minority of infected tumor cells in vivo (20,47,99). While lytic infection may not be compatible with transformation, lytic gene products are thought to have important paracrine effects on adjacent latently infected or uninfected cells that are vital for lesion formation. In addition, viral gene expression patterns may be tumor or stage specific, and some early lytic genes may be expressed for an extended period of time in the absence of a complete replication cycle. Thus, KSHV encodes an arsenal of proteins that could conceivably induce and/or maintain KS lesions. Despite this recognition, the mechanisms of virus-induced oncogenesis remain unclear.In vitro studies with KSHV were initially performed using PEL cell lines established from PEL tumors that comprise a rare form of AIDS-associated B-cell lymphoma (7,14,22 (75,86,93,103) and have provided a source of infectious KSHV for infection of other cell types. In addition, DNA array analysis of PEL cells has been used to map the transcription program of KSHV (57,78). Endothelial cells are, however, a more relevant cell type in which to study KS pathogenesis, since they are the likely precursors of KS spindle cells (11,90,91,96). Interestingly though, cells cultured from KS tumors do not maintain the KSHV genome (3, 4), and the early consequences of cellular transformation are inaccessible via the study of fully transformed tumor cells. In vitro endothelial cell models of KSHV infection thus provide the most useful systems with which to dissect the role of KSHV in KS spindle cell development and growth, but development of such systems has proved to be challenging. To date several endothelial-based models of KSHV...