Division and apoptosis of E2f-deficient retinal progenitors

The activating E2f transcription factors (E2f1, E2f2 and E2f3) induce transcription and are widely viewed as essential positive cell cycle regulators. Indeed, they drive cells out of quiescence, and the 'cancer cell cycle' in Rb1 null cells is E2f-dependent(1,2). Absence of activating E2fs in flies or mammalian fibroblasts causes cell cycle arrest(3,4), but this block is alleviated by removing repressive E2f or the tumour suppressor p53, respectively(5-7). Thus, whether activating E2fs are indispensable for normal division is an area of debate(1). Activating E2fs are also well known pro-apoptotic factors, providing a defence against oncogenesis(8), yet E2f1 can limit irradiation-induced apoptosis(9,10). In flies this occurs through repression of hid (also called Wrinkled; Smac/Diablo in mammals). However, in mammals the mechanism is unclear because Smac/Diablo is induced, not repressed, by E2f1(11), and in keratinocytes survival is promoted indirectly through induction of DNA repair targets(12). Thus, a direct pro-survival function for E2f1-3 and/or its relevance beyond irradiation has not been established. To address E2f1-3 function in normal cells in vivo we focused on the mouse retina, which is a relatively simple central nervous system component that can be manipulated genetically without compromising viability and has provided considerable insight into development and cancer(2,13). Here we show that unlike fibroblasts, E2f1-3 null retinal progenitor cells or activated Muller glia can divide. We attribute this effect to functional interchangeability with Mycn. However, loss of activating E2fs caused down-regulation of the p53 deacetylase Sirt1, p53 hyperacetylation and elevated apoptosis, establishing a novel E2f-Sirt1-p53 survival axis in vivo. Thus, activating E2fs are not universally required for normal mammalian cell division, but have an unexpected pro-survival role in development.