Deregulation of the Cell Cycle by EBV Nuclear Antigens EBNA3A and EBNA3C

Abstract

Cyclin-dependent kinase inhibitor p16INK4A is an important tumour suppressor and inducer of cellular senescence often inactivated during the development of cancer. I investigated the mechanism by which EBV latency-associated nuclear antigens EBNA3A and EBNA3C repress p16INK4A expression. Using lymphoblastoid cell lines (LCL) expressing a conditional EBNA3C, I demonstrate that EBNA3C inactivation resets the epigenetic status of p16INK4A to permit transcriptional activation: the polycomb-associated repressive H3K27me3 histone modification is substantially reduced, while the activation-related mark H3K4me3 is modestly increased. Activation of EBNA3C reverses the distribution of these epigenetic marks, represses p16INK4A transcription and allows proliferation. LCL lacking EBNA3A express relatively high levels of p16INK4A and have a similar pattern of histone modifications on p16INK4A as produced by the inactivation of EBNA3C. Since binding to the co-repressor of transcription CtBP was linked to the oncogenic activity of EBNA3C and EBNA3A, LCL with viruses encoding EBNA3C- and/or EBNA3A-mutants that no longer bind CtBP were established. These novel LCL revealed that the epigenetic repression of p16INK4A requires the interaction of both EBNA3C and EBNA3A with CtBP. Epigenetic repression of p16INK4A by latent EBV may facilitate p16INK4A DNA methylation during lymphomagenesis. Furthermore, by transforming the peripheral blood lymphocytes (PBL) from an individual homozygous for a deletion in CDKN2A locus with recombinant EBV viruses expressing conditional EBNA3C, we developed a system that allows inactivation of EBNA3C in LCL lacking functional p16INK4A protein (p16-null LCL 3CHT). EBNA3C inactivation has no impact on the proliferation rate of p16-null LCL, proving that the repression of p16INK4A is the main function of EBNA3C in EBV-driven LCL proliferation. The p16INK4A locus is epigenetically modified by EBNA3C despite the absence of functional p16INK4A protein. Since the selection pressure based on faster outgrowth of advantageously modified subset of cells is removed, the gradual and relatively slow kinetics of H3K27me3 restoration at p16INK4A following EBNA3C reactivation in p16-null LCL 3CHT seems to be genuinely related to the mechanism of EBNA3C-mediated p16INK4A regulation. The p16-null LCL 3CHT system further allows distinguishing genes regulated specifically by EBNA3C, rather than as a consequence of activation of p16INK4A/Rb/E2F1 axis. Lastly, new cellular targets of EBNA3C and/or EBNA3A from the group of microRNAs are identified in this work. Most notably, both EBNA3C and EBNA3A are shown to repress the tumour supressor miR-143/145 cluster and their precursor long non-coding RNAs in LCL.