An endothelial cell line infected by Kaposi's sarcoma-associated herpes virus (KSHV) allows the investigation of Kaposi's sarcoma and the validation of novel viral inhibitors in vitro and in vivo.

Kaposi's sarcoma-associated herpesvirus (KSHV) is the etiological agent of Kaposi's sarcoma (KS), a tumor of endothelial origin predominantly affecting immunosuppressed individuals. Up to date, vaccines and targeted therapies are not available. Screening and identification of anti-viral compounds are compromised by the lack of scalable cell culture systems reflecting properties of virus-transformed cells in patients. Further, the strict specificity of the virus for humans limits the development of in vivo models. In this study, we exploited a conditionally immortalized human endothelial cell line for establishment of in vitro 2D and 3D KSHV latency models and the generation of KS-like xenograft tumors in mice. Importantly, the invasive properties and tumor formation could be completely reverted by purging KSHV from the cells, confirming that tumor formation is dependent on the continued presence of KSHV, rather than being a consequence of irreversible transformation of the infected cells. Upon testing a library of 260 natural metabolites, we selected the compounds that induced viral loss or reduced the invasiveness of infected cells in 2D and 3D endothelial cell culture systems. The efficacy of selected compounds against KSHV-induced tumor formation was verified in the xenograft model. Together, this study shows that the combined use of anti-viral and anti-tumor assays based on the same cell line is predictive for tumor reduction in vivo and therefore allows faithful selection of novel drug candidates against Kaposi's sarcoma. KEY MESSAGES: Novel 2D, 3D, and xenograft mouse models mimic the consequences of KSHV infection. KSHV-induced tumorigenesis can be reverted upon purging the cells from the virus. A 3D invasiveness assay is predictive for tumor reduction in vivo. Chondramid B, epothilone B, and pretubulysin D diminish KS-like lesions in vivo.