We showed that Epstein-Barr virus EBNA3A and EBNA3C act as oncoproteins but EBNA3B behaves as a tumour suppressor, the first described in a tumour virus. Our reverse genetic analysis revealed that EBNA3A, 3B and 3C together regulate the expression of specific host genes in B cells, often repressing transcription. This repression involves recruitment of polycomb group (PcG) proteins and the epigenetic mark of repression H3K27me3. Target genes include those encoding the inhibitor of proliferation p16INK4a and pro-apoptotic factor BIM, both of which are repressed in a similar manner in stem cell maintenance and are often silenced in cancer. We have thus proposed that EBV epigenetically reprograms B cells in viral persistence and in B cell lymphomas. The regulation of PcG-mediated repression linked to EBV-associated cancers gives an exceptional opportunity to investigate, in a physiological context, principles of epigenetic silencing. These are likely to have wide application to cancer biology and epigenetic gene regulation by persistent microorganisms. Our key goals are: Detailed biochemical mechanisms for interactions between the EBNA3s and cellular factors responsible for this EBV-mediated reprogramming. Comprehensive genome-wide analyses establishing the extent of EBV-induced reprogramming via the PcG system and the identities of all genes specifically targeted by EBNA3s. To use EBNA3 proteins as tools to understand poorly defined processes in PcG-mediated repression such as the sequence of events in recruitment of complexes to specific target genes. To use our animal model to determine how these processes contribute to EBV persistence, B cell lymphomagenesis and tumour immunosurveillance.