Visualizing Dynamic E2F-Mediated Repression In Vivo

Although many E2F target genes have been identified recently, very little isknown about how any single E2F site controls the expression of an E2Ftarget gene in vivo. To test the requirement for a single E2F site invivo and to learn how E2F-mediated repression is regulated duringdevelopment and tumorigenesis, we have constructed a novel series ofwild-type and mutantRb promoter-LacZ transgenicreporter lines that allow us to visualize the activity of a crucial E2Ftarget in vivo, the retinoblastoma tumor suppressor gene (Rb ).Two mutantRb promoter-LacZ constructs were used toevaluate the importance of a single E2F site or a nearby activator(Sp1/Ets) site that is found mutated in low-penetrance retinoblastomas.The activity of the wild-typeRb promoter is dynamic, varyingspatially and temporally within the developing nervous system. Whileloss of the activator site silences theRb promoter, loss ofthe E2F site stimulates its activity in the neocortex, retina, andtrigeminal ganglion. Surprisingly, E2F-mediated repression ofRb does not act globally or in a static manner but, instead,is a highly dynamic process in vivo. Using neocortical extracts, wedetected GA-binding protein α (GABPα, an Ets family member) bound to the activator site and both E2F1 and E2F4 bound to the repressor site of theRb promoter in vitro. Additionally, we detected binding of both E2F1 and E2F4 to theRb promoter invivo using chromatin immunoprecipitation analysis on embryonic day 13.5brain. Unexpectedly, we detect no evidence forRb promoterautoregulation in neuroendocrine tumors fromRb +/ −;RbP-LacZ mice that undergo loss of heterozygosity at theRb locus, in contrast to the situation in humanretinoblastomas where highRB mRNA levels are found. Insummary, this study provides the first demonstration that loss of anE2F site is critical for target gene repression in vivo and underscoresthe complexity of theRb and E2F family network invivo.