ORIGINAL ARTICLE p53 directs focused genomic responses in Drosophila F Akdemir 1 , A Christich 1 , N Sogame 1 , J Chapo 2 and JM Abrams Department of Cell Biology, UT Southwestern Medical Center, Dallas, TX, USA p53 is a fundamental determinant of cancer susceptibility and other age-related pathologies. Similar to mammalian counterparts, Drosophila p53 integrates stress signals and elicits apoptotic responses that maintain genomic stabi- lity. To illuminate core-adaptive functions controlled by this gene family, we examined the Drosophila p53 regulatory network at a genomic scale. In development, the absence of p53 impacted constitutive expression for a surprisingly broad scope of genes. By contrast, stimulus- dependent responses governed by Drosophila p53 were limited in scope. The vast majority of stress responders were induced and p53 dependent (RIPD) genes. The signa- ture set of 29 ‘high stringency’ RIPD genes identified here were enriched for intronless loci, with a non-uniform distribution that includes a recently evolved cluster unique to Drosophila melanogaster. Two RIPD genes, with known and unknown biochemical activities, were func- tionally examined. One RIPD gene, designated XRP1, maintains genome stability after genotoxic challenge and prevents cell proliferation upon induced expression. A second gene, RnrL, is an apoptogenic effector required for caspase activation in a model of p53-dependent killing. Together, these studies identify ancient and convergent features of the p53 regulatory network. Oncogene (2007) 26, 5184–5193; doi:10.1038/sj.onc.1210328; published online 19 February 2007 Keywords: apoptosis; p53; radiation; Drosophila; cell death Introduction p53 is a critical tumor suppressor gene found mutated or altered in most human cancers (reviewed in Sharpless and DePinho, 2002). Members of the p53 family are encoded in the genomes of flies and worms and, like their mammalian counterparts, these invertebrate homologs also limit cell growth in response to genotoxic stress and promote cell death in numerous pathologic models (Vousden and Prives, 2005). A central focus of p53 research seeks to understand how this protein family integrates pathologic stimuli and functions to specify adaptive (or maladaptive) cellular responses. In mammals, p53 restrains cell growth by halting prolifera- tion and/or inducing apoptosis but, in Drosophila and in Caenorhabditis elegans, the p53 ortholog influences apoptosis without exerting significant effects upon cell cycle progression (reviewed in Lu and Abrams, 2006). A primary mode of p53 action operates to regulate target genes at the level of transcription and, hence, a compelling focus of cancer research is driven by efforts to understand the downstream targets of p53 (reviewed in Lu and Abrams (2006); and see Supplementary Table S5). In many cases, regulated expression of downstream genes is context-specific and whether there exists a ‘generic’ p53program that drives apoptosis in all cells is not known. Many genome-wide searches for p53- responsive genes have also been reported, but the extent to which these experiments clarify authentic targets is not clear as most have examined dominant conse- quences of ectopic overexpression in (1) otherwise unstressed contexts and/or (2) ‘transformed’ cells in culture. Therefore, interpretations derived from these studies must consider implications associated with ‘off- target’ effects. Studies in Drosophila illuminated several inhibitor of apoptosis proteins antagonists, reaper (rpr), sickle (skl) and head involution defective (hid), as important targets of p53 action in vivo (Lee et al., 2003b; Sogame et al., 2003; Brodsky etal., 2004). These findings and a recently implicated link to the Hippo pathway (Colombani etal., 2006) have relevance to human cancers as genes with similar activities are also p53 targets in mammals (Jin et al., 2003). Hence, in both systems, p53 can link damage signals to apoptosis by engaging proteins that de-repress caspase activity. To initiate a more complete picture of ancestral functions governed by p53, we profiled expression as a function of p53 status in the Drosophila model. Our rationale involved comparative analyses of radiation-responsive genes in wild-type and p53 mutants together with computed determinations of p53-binding sites and functional assays in cultured cells and in the animal. In contrast to studies that rely upon forced p53 expression, here we apply array technologies to a context where p53 function has been genetically removed. Therefore, an important strength of our approach derives from a focus on stimulus-dependent responses in a loss-of-p53 background. Received 14 September 2006; revised 28 November 2006; accepted 9 January 2007; published online 19 February 2007 Correspondence: Professor J Abrams, Department of Cell Biology, UT Southwestern Medical Center, Dallas, TX 75390, USA. E-mail: John.Abrams@utsouthwestern.edu 1 These three authors contributed equally to this work. 2 Current address: Myogen Inc., Westminster, CO 80021,USA. Oncogene (2007) 26, 5184–5193 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc