The biology of Kaposi sarcoma is poorly understood because the dominant cell type in Kaposi sarcoma lesions is not known 1-4. We show by gene expression microarrays that neoplastic cells of Kaposi sarcoma are closely related to lymphatic endothelial cells (LECs) and that Kaposi sarcoma herpesvirus (KSHV) 5,6 infects both LECs and blood vascular endothelial cells (BECs) in vitro. The gene expression microarray profiles of infected LECs and BECs show that KSHV induces transcriptional reprogramming of both cell types. The lymphangiogenic molecules VEGF-D and angiopoietin-2 were elevated in the plasma of individuals with acquired immune deficiency syndrome and Kaposi sarcoma. These data show that the gene expression profile of Kaposi sarcoma resembles that of LECs, that KSHV induces a transcriptional drift in both LECs and BECs and that lymphangiogenic molecules are involved in the pathogenesis of Kaposi sarcoma. The cellular origin of the spindle cells of Kaposi sarcoma lesions is poorly defined 3,7. Kaposi sarcoma spindle cells express endothelial cell markers but also have features of other cell lineages, including fibroblasts, macrophages and smooth muscle cells 1-3. Kaposi sarcoma could be a tumor originating from LECs, as spindle cells ubiquitously express VEGFR-3 and podoplanin and stain with the antibody D2-40 recognizing LECs 4,8. But these markers can also be expressed on angiogenic blood vessels, or on other cell types 9. Furthermore, some BEC markers (e.g., CD34) are expressed in all Kaposi sarcoma spindle cells 1. KSHV is the infectious cause of Kaposi sarcoma 5,6. In vitro, KSHV can infect both micro-and macrovascular endothelial cells, and these cells are useful to study the role of KSHV in the pathogenesis of Kaposi sarcoma 10-12 .

Kaposi's sarcoma (KS)-associated herpesvirus (KSHV) is the infectious cause of KS and is also linked to the pathogenesis of certain lymphoproliferations (4,14). It is proposed that KSHV latent proteins are directly involved in modulating signal transduction pathways and cellular circuits leading to uncontrolled cell proliferation (2).At the far right-hand end of the KSHV genome, open reading frame (ORF) K15 encodes a putative transmembrane protein in the same genomic location as the Epstein-Barr virus (EBV) latent membrane protein 2A (LMP2A) (5,7,12,39). K15 resembles LMP2A not only in genomic location but also in its splicing pattern and predicted protein structure. Two highly divergent forms of K15 have been identified: the predominant (P) and minor (M) forms (5, 12, 39). These two alleles possess only 33% amino acid identity yet retain 12 membrane-spanning domains and a putative cytoplasmic signal-transducing carboxyl terminus (C terminus) (5). The C terminus of K15 has potential signaling motifs, including Src homology 2 and 3 binding domains (SH2-B and SH3-B, respectively) (12). A CD8-K15 C-terminal chimeric protein was shown to be constitutively tyrosine phosphorylated within the SH2-B motif (5). Like LMP2A, this CD8-K15 chimeric protein modulates B-cell receptor (BCR) signal transduction. The mechanism(s) of signal transduction is unknown but appears to be distinct from that of LMP2A and does not involve intracellular free calcium mobilization (5).In addition, the C terminus of K15 has sequences similar to those found in EBV LMP1, including a putative tumor necrosis factor receptor-associated factor (TRAF) binding site. K15 therefore appears to be a hybrid of a distant evolutionary relative of both EBV LMP1 and LMP2A (13). The putative C terminus of K15 has been shown to interact with the TRAFs (12), and we have also shown that K15 can indeed activate NF-B via this putative TRAF binding site (unpublished data). By way of activating NF-B, LMP1 of EBV plays an essential role in EBV-induced transformation of B lymphocytes (3,16,21). NF-B activation also appears to be essential for the proliferation potential of KSHV positive primary effusion lymphoma (PEL) cells (22), but whether all of this NF-B activity in PEL cells is due to K15 expression is not yet known.Although K15 mRNA has been demonstrated in PEL cells (5, 12, 39), it is not known whether the K15 protein is actually expressed in latently infected tumor cells. The size of endogenous protein, its exact subcellular localization, and its cellular binding partners have not previously been determined.We generated a monoclonal antibody (MAb) against K15 and show here that when K15 cDNA is ectopically expressed we detect the predicted 50-kDa form as well as a series of smaller proteolytically cleaved forms, of which the 35-and 23-kDa species are predominant. Deletion of the initiator AUG of the K15 ORF abolished protein expression, suggesting that the 50-kDa form of K15 is a precursor which is subsequently proteolytically processed into smaller species. We ...

No abstract