Most cancers harbor mutations that directly or indirectly compromise the function of the retinoblastoma tumor suppressor protein (pRB). The best studied function of pRB relates to its ability to bind to members of the E2F transcription factor family and for such pRB-E2F complexes to actively repress the transcription of E2F-responsive promoters. We discovered that some pRB variants that cannot bind to E2F nonetheless promote senescence or differentiation when reintroduced into RB1-/-tumor cells. Using such a pRB variant as a 2-hybrid bait, we recovered RBP2 (also called JARID1A or KDM5A) and confirmed that endogenous pRB-RBP2 complexes can be detected in chromatin- enriched cell extracts. Moreover, we showed that the ability of pRB variants to promote senescence and differentiation mirrors their ability to bind to RBP2 and that RBP2 siRNA is sufficient to promote senescence or differentiation in RB1-/- tumor cells. More recently we showed that RBP2 is an H3K4 demethylase and our preliminary data suggest that loss of RBP2 suppresses pRB-defective tumor growth in vivo as well the growth of tumors lacking the tumor suppressor MEN1, which is a component of an H3K4 methyltransferase complex. In aims 1 and 2 we will ask whether loss of RBP2 histone demethylase activity suppresses tumor growth in vivo. RBP2 belongs to a superfamily of 2-oxoglutarate-dependent dioxygenases. Aims 1 and 2 will be complemented in Aim 3 by an unbiased shRNA screen for 2-oxoglutarate-dependent dioxygenases (and other selected histone modifiers) that, when inhibited, impair the fitness of pRB-defective tumors cells and Aim 4 will develop cell-based assays for inhibitors of such enzymes, using RBP2 as the prototype.