Background:  Kaposi’s sarcoma-associated herpesvirus (KSHV) is a tumorigenic virus that establishes lifelong infection in the human host by infecting B cells and expresses viral proteins capable of modulating B cell signaling and survival. KSHV infection in an immunocompromised host results in several tumors, including 2 B cell dyscrasias (multicentric Castleman’s disease, and primary effusion lymphoma). To date, the only clues to the identity of the human B cell subsets that may represent the earliest targets during KSHV transmission are inferences based on the phenotypic characterization of KSHV⁺ lymphomas. Another hint, however, may come from Chagas’ et al. (2006) histological study of human tonsils, which found KSHV⁺ cells localized to the region beneath the tonsillar crypts. Since this area is populated by a unique subset of B cells (subepithelial B cells) and epidemiologic data support salivary transmission of KSHV, we hypothesized that subepithelial B cells may represent an initial entry point, as well as a long-term reservoir for the virus. The surface marker phenotype of multicentric Castleman’s disease is consistent with this hypothesis. Identification and characterization of B cell subsets infected by KSHV may shed light into the mechanisms of KSHV transmission as well as the genesis of KSHV-induced tumors in patients with AIDS or other immune deficiencies.

Methods:  We exposed cultures of purified tonsillar B cells to KSHV and then used multispectral imaging fluorescence cytometry (Imagestream, Amnis), a high through-put method for analyzing both immunofluorescence and morphological parameters to analyzed the surface marker phenotype of cells expressing the major latent viral protein.

Results:  We found that between 1% and 16% of human tonsillar B cells became infected upon exposure to KSHV, that these infected B cells were immunoglobulin (Ig) M^hi, IgD^lo, and CD5⁺ and that CD5 expression increased between 48 and 72 hours post inoculation. Furthermore, KSHV-infected cells were restricted to those expressing the lambda immunoglobulin light chain.

Conclusions:  The phenotype of B cells infected with KSHV in vitro corresponds to activated (CD5⁺) subepithelial (IgM^hi, IgD^lo) B cells. Additionally, the lambda-restriction of KSHV-infected subepithelial B cells mirrors this phenotype in multicentric Castleman’s disease. These data suggest that KSHV preferentially infects subepithelial B cells ex vivo, implying that there may be both functional as well as anatomic aspects of this population that favor infection.