Cancer cells have abnormal gene expression profiles; however, to what degree these are chaotic or driven by structured gene regulatory networks is often not known. Here we studied a model of Ras-driven invasive tumorigenesis in Drosophila epithelial tissues and combined in vivo genetics with next-generation sequencing and computational modeling to decipher the regulatory logic of tumor cells. Surprisingly, we discovered that the bulk of the tumor-specific gene expression is controlled by an ectopic network of a few transcription factors that are overexpressed and/or hyperactivated in tumor cells. These factors are Stat, AP-1, the bHLH proteins Myc and AP-4, the nuclear hormone receptor Ftz-f1, the nuclear receptor coactivator Taiman/SRC3, and Mef2. Notably, many of these transcription factors also are hyperactivated in human tumors. Bioinformatic analysis predicted that these factors directly regulate the majority of the tumor-specific gene expression, that they are interconnected by extensive cross-regulation, and that they show a high degree of co-regulation of target genes. Indeed, the factors of this network were required in multiple epithelia for tumor growth and invasiveness, and knockdown of several factors caused a reversion of the tumor-specific expression profile but had no observable effect on normal tissues. We further found that the Hippo pathway effector Yorkie was strongly activated in tumor cells and initiated cellular reprogramming by activating several transcription factors of this network. Thus, modeling regulatory networks identified an ectopic and ordered network of master regulators that control a large part of tumor cell-specific gene expression.

The Hippo tumor-suppressor pathway regulates organ growth, cell proliferation, and stem cell biology. Defects in Hippo signaling and hyperactivation of its downstream effectors—Yorkie (Yki) in Drosophila and YAP/TAZ in mammals—result in progenitor cell expansion and overgrowth of multiple organs and contribute to cancer development. Deciphering the mechanisms that regulate the activity of the Hippo pathway is key to understanding its function and for therapeutic targeting. However, although the Hippo kinase cascade and several other upstream inputs have been identified, the mechanisms that regulate Yki/YAP/TAZ activity are still incompletely understood. To identify new regulators of Yki activity, we screened in Drosophila for suppressors of tissue overgrowth and Yki activation caused by overexpression of atypical protein kinase C (aPKC), a member of the apical cell polarity complex. In this screen, we identified mutations in the heterogeneous nuclear ribonucleoprotein Hrb27C that strongly suppressed the tissue defects induced by ectopic expression of aPKC. Hrb27C was required for aPKC-induced tissue growth and Yki target gene expression but did not affect general gene expression. Genetic and biochemical experiments showed that Hrb27C affects Yki phosphorylation. Other RNA-binding proteins known to interact with Hrb27C for mRNA transport in oocytes were also required for normal Yki activity, although they suppressed Yki output. Based on the known functions of Hrb27C, we conclude that Hrb27C-mediated control of mRNA splicing, localization, or translation is essential for coordinated activity of the Hippo pathway.

RNA binding proteins and messenger RNAs (mRNAs) assemble into ribonucleoprotein granules that regulate mRNA trafficking, local translation, and turnover. The dysregulation of RNA-protein condensation disturbs synaptic plasticity and neuron survival and has been widely associated with human neurological disease. Neuronal granules are thought to condense around particular proteins that dictate the identity and composition of each granule type. Here, we show in Drosophila that a previously uncharacterized long noncoding RNA, mimi , is required to scaffold large neuronal granules in the adult nervous system. Neuronal ELAV-like proteins directly bind mimi and mediate granule assembly, while Staufen maintains condensate integrity. mimi granules contain mRNAs and proteins involved in synaptic processes; granule loss in mimi mutant flies impairs nervous system maturity and neuropeptide-mediated signaling and causes phenotypes of neurodegeneration. Our work reports an architectural RNA for a neuronal granule and provides a handle to interrogate functions of a condensate independently of those of its constituent proteins.

Like many other animals, insects are capable of returning to previously visited locations using path integration. Recently, Drosophila has been added to the list of insects thought capable of this navigational memory. Existing experimental evidence, however, has a potential confound. Here we show that pheromones deposited at the site of reward might enable flies to find previously rewarding locations even without memory. Thus, we designed an experiment to determine if flies can use path integration memory despite potential pheromonal cues by displacing the flies shortly after an optogenetic reward. We found that rewarded flies returned to the location predicted by a memory-based model. Several analyses are consistent with path integration as the mechanism by which flies returned to the reward. We conclude that while pheromones may often be important in fly navigation and must be carefully controlled in future experiments, Drosophila may indeed be capable of performing path integration.