dE2F1 and Yki cooperate to drive inappropriate cell divisions. (A–D) Images of third instar larval eye discs, where the posterior is positioned to the right and the morphegenetic furrow (MF) is marked by a white arrowhead. In all images, differentiated cells are shown by the presence of ELAV (blue) and proliferating cells are marked by the presence of phosphorylated Histone-H3 (pH3) (green). Ectopic expression of UAS-de2f1, UAS-yki^S168A, or the two together was driven exclusively in the posterior compartment by the GMR-Gal4 driver. (A) Proliferation is absent from the posterior compartment of wild-type eye discs. (B,C) When de2f1 or yki is overexpressed, inappropriate cell divisions marked by pH3 (green) can be detected in the posterior. (D) Overexpression of both de2f1 and yki together leads to massive cell division in the posterior compartment. (E) Upon quantification (for details, the see Materials and Methods), there is an apparent cooperation between dE2F1 and Yki in driving inappropriate cell divisions (P-value < 0.00003 compared with either dE2F1 or Yki overexpression alone). Error bars are representative of standard deviation from the mean in each experiment.

dE2F/RBF and Yki/Sd complexes can bind directly to promoters of shared targets. Schematics of the promoter regions of DNA pol ɛ (A), dDP (B), Mcm3 (C), and Mcm10 (D). The depicted regions are each a total of 1 kb in length, with the grayscale bar representing 0.1 kb in length, andthe transcription start site (TSS) for each promoter is designated by a black arrow. Fragments from within these four regions were used in cloning the reporters described above (for details, see the Supplemental Material). In all of the schematics, the predicted Sd-binding sites within each promoter region are shown by the open triangles on top, and the predicted dE2F-binding sites are shown by closed triangles on the bottom. The closed bars represent regions (∼0.15 kb) amplified in real-time qPCR analysis following ChIP experiments using RBF, Dp, and dE2F1 antibodies, as well as a Flag antibody to isolate Sd/DNA complexes. All closed bars represent amplicons from unique primer sets. ChIP was performed using antibodies specific for dDP, RBF, or dE2F1. An antibody recognizing the MYC epitope served as a negative control. To determine occupancy by Sd, S2R⁺ cells were transfected with Flag-Sd and HA-Yorkie. ChIP was performed with an antibody specific for the Flag epitope. Untransfected cells served as a negative control. Real-time qPCR analysis was done on the recovered chromatin samples, and the total amount of DNA precipitated with each antibody was quantified. RpP0 served as the negative control gene for each experiment. Fold enrichment was calculated by dividing chromatin enriched from transfected cells by untransfected cells for each region specified. Error bars are representative of a standard deviation from the mean. All differences were statistically significant, with a P-value of <0.08.

An rbf mutation partially relieves the effect of the loss of de2f1 on yki-driven proliferation in vivo. (A) Loss of de2f1 prevents yki-induced hyperplasia. The MARCM system is used to overexpress yki in both a wild-type and de2f1^−/− background. Tissue overexpressing yki is marked by the presence of intense GFP. (B) Inactivation of the Hippo pathway through a mutation in wts, sav, or mats leads to hyperplasia of the adult eye due to severe proliferation and inhibition of apoptosis. Concomitant loss of de2f1 in the background of Hippo pathway inactivation leads to a dramatic reduction in mutant tissue representation. Mosaic tissue was generated by the FLP/FRT system. Wild-type tissue is identified by the presence of red pigment. Mutant tissue is identified by the absence of red pigment. (C–D′) Clonal twin spot analysis done in the third instar larval wing disc of either de2f1 wts double mutant (C,C′) or rbf de2f1 wts triple mutant (D,D′) tissue compared with wild-type tissue. Nineteen total clonal twin spots were analyzed for each comparison, and the total number of cells for each genotype was then counted. Wild-type tissue is marked by two copies of GFP and is outlined in white in C′ and D′. Mutant tissue was represented by the absence of GFP and is outlined in yellow in C′ and D′. The histograms in C and D represent the distribution of the total number of cells detected in each clonal twin spot. Wild-type samples are represented by the open bars, and the mutant samples are represented by yellow bars. Differences between each mutant cell population and the corresponding wild-type sample were statistically significant, with a P-value of <0.002. (E) Depletion of RBF partially restores reduced level of expression of reporters in dE2F1-depleted cells. Cells were incubated with nonspecific (White), Wts, or a combination of Wts, RBF, and RBF2 dsRNAs for 4 d to deplete the corresponding proteins. On day 4, cells were incubated with dE2F1 dsRNA for 24 h. Following 24 h, cells were transfected with PCNA-luc, Mcm3-luc, and Mcm10-luc reporters. Depletion of dE2F1 greatly reduces endogenous activity from the PCNA-luc, Mcm3-luc, and Mcm10-luc reporters. Codepletion of Warts is unable to rescue this reduced activity. However, upon depletion of Warts, RBF, and RBF2 together, activity from each reporter is increased (P-value < 0.005).

RBF and Hippo pathways cooperatively regulate DNA replication and cell cycle pathway genes. Three-way Venn diagrams of differentially expressed (DE) genes from rbf/wild type, wts/wild type, and rbf wts (double mutant [DM])/wild type. The three-way Venn diagrams represent DE genes (compared with wild type) when there is a log₂ fold change (FC) >0.5 (∼1.4-fold change). (A) On the left, up-regulated DE genes in rbf/wild type, wts/wild type, and DM/wild type. Using a log₂FC > 0.5, 55 DE genes are up-regulated only following inactivation of both the RBF and Hippo pathways. On the right, down-regulated DE genes in rbf/wild type, warts/wild type, and DM/wild type. Using a log₂FC > 0.5, 90 DE genes are down-regulated only following inactivation of both the RBF and Hippo pathways. Uniquely DE genes found in the rbf wts double mutant tissue represent a synergy between the two mutant alleles on gene expression. (B) Matrix of GOBP within each mutant genotype (rbf, wts, and rbf wts). Enrichment of genes up-regulated in each genotype can be seen for cell cycle and DNA replication. (C) GOBP enrichment analysis for DE genes found to be either uniquely up-regulated in the rbf wts double mutant or commonly within all three mutant genotypes (rbf, wts, and rbf wts). For all matrices in B and C, colors are values of statistical significance (see scale bar) for enrichment of each process, with red being the most significant and gray indicating no significant enrichment. (D) Real-time qPCR analysis of RNA isolated from wild-type tissue or rbf, wts, or rbf wts mutant tissues. Classical target genes from each pathway, such as expanded (Hippo pathway) (P-value < 0.1) or BubR1 (RBF), show increased expression following inactivation of both pathways. Error bars represent standard deviation from the mean.

dE2F1 and Yki/Sd synergistically activate common target genes. (A) Statistical analysis revealed that a statistically significant number of DE up-regulated genes found in the rbf and rbf wts mutant tissues have both Sd- and dE2F1-binding sites within the promoters. GOBP enrichment analysis of these DE up-regulated genes found a gene expression signature similar to that of the rbf and rbf wts mutants (enrichment for processes involved in DNA replication). (B) An enrichment of genes involved in DNA replication and cell cycle progression pathways was found, in the DE up-regulated genes from the rbf wts double mutant, to have both Sd- and dE2F1-binding sites. (C–F) Yki/Sd and dE2F1 cooperate in activation of the reporter in transient transfection experiments in S2R⁺ cells. (C) Cells were incubated with nonspecific (NS), RBF/RBF2, and Hippo (Hpo) dsRNAs for 4 d to deplete the corresponding proteins. On day 4, the E2F-Sd-Luc reporter was transfected into the depleted cells with pIE7-Flag-Sd- and pAc-HA-Yki-expressing plasmids, and the luciferase activity was measured 2 d later to determine the level of the endogenous E2F activity in these cells. The pAc-Renilla plasmid was cotransfected to normalize for transfection efficiency. Results depict the mean of three experiments. (D) S2R⁺ cells were transfected with the Sd-E2F-luc reporter and a combination of pIE7-Flag-Sd-, pAc-HA-Yki-, pIE4-Myc-E2F1-, and pIE4-HA-RBF-expressing plasmids. Yki/Sd and dE2F1 exerted a cooperative effect in activation of the reporter, while coexpression of RBF negated such cooperation. The pAc-Renilla plasmid was cotransfected to normalize for transfection efficiency. Results depict the mean of three experiments. (E,F) S2R⁺ cells were transfected with luciferase reporters containing fragments of endogenous promoters from genes that are predicted dual targets of Sd and dE2F and a combination of pIE7-Flag-Sd-, pAc-HA-Yki-, pIE4-Myc-E2F1-, and pIE4-HA-RBF-expressing plasmids. Yki/Sd and dE2F1 exerted a cooperative effect in activation of the reporters, while coexpression of RBF negated such cooperation. For Mcm3-luc, dDP-luc, Dachs-luc, Ex-luc, and DNA Pol ɛ- luc, this cooperative effect in activation was a statistically significant (P-value < 0.06) increase in activation seen when the same amount of dE2F1 or Yki/Sd were expressed singularly. The difference seen in activation for Mcm2-luc, Mcm10-luc, cdc2c-luc, and CycB3-luc was more modest (P-value < 0.105). The Sd-E2F-luc reporter was used as a positive control and is denoted by a red asterisk. The pAc-Renilla plasmid was cotransfected to normalize for transfection efficiency. Results depict the mean of three experiments. (G) S2R⁺ cells were treated with nonspecific (NS), RBF/RBF2, Hpo/Warts, or RBF/RBF2 + Hpo/Warts dsRNAs for 4 d to deplete the corresponding proteins. On day 4, RNA was recovered from each treatment, and real-time qPCR analysis was performed on the corresponding cDNAs. Expression of all genes tested, with the exception of Ex, in the RBF/RBF2 + Hpo/Warts-codepleted cells was statistically elevated (P-value < 0.07) when compared with cells treated with a nonspecific dsRNA. Results depict mean of three experiments.

Distinct regulation of gene expression through activation by Yki/Sd and dE2F1. (A–C) Cells were incubated with nonspecific (White), dE2F1, and a combination of both Yki and Sd dsRNAs for 24 h to deplete the corresponding proteins. Following 24 h, the MCM3-luc, dDP-luc, and DNA pol ɛ-luc reporters were transfected into the depleted cells with pIE4-HA-RBF-, pIE4-HA-RBF2-, and pIE-Myc-E2F2-expressing plasmids, and the luciferase activity was measured 2 d later. The pAc-Renilla plasmid was cotransfected to normalize for transfection efficiency. Depletion of Yki/Sd reduced the expression of dDP-luc and DNA pol ɛ-luc reporters (P-value < 0.07). Depletion of dE2F1 reduced the expression of the three tested reporters between 50% and 80% (P-value < 0.06). Results depict the mean of three experiments. Depletion of dE2F1 mildly enhanced dE2F2/RBF-mediated repression of the dDP-luc reporter (P-value < 0.06). Depletion of Yki/Sd mildly enhanced dE2F2/RBF-mediated repression of the Mcm3-luc reporter (P-value < 0.05). (D) Yki/Sd failed to activate the reporters in dE2F1-depleted cells to the same extent as in control treated cells. Nonspecific (White) or dE2F1 proteins were depleted for 24 h, and then cells were transfected with the three reporters in A–C and either empty vector or pIE7-Flag-Sd- and pAc-HA-Yki-expressing plasmids. (E) dE2F2 represses the PCNA-luc reporter in control treated and dE2F1-depleted cells. Following 24 h of depletion, the PCNA-Luc reporter was transfected together with pIE4-HA-RBF-, pIE4-HA-RBF2-, and pIE-Myc-E2F2-expressing plasmids. The pAc-Renilla plasmid was cotransfected to normalize for transfection efficiency. (F) dE2F2 binds at the promoters of dDP, Mcm10, DNA pol ɛ, and Mcm3. ChIP was performed as described in Figure 4. (G) Inactivation of the Hippo pathway does not disrupt interaction between RBF and dE2F2. Immunoprecipitation-Western was performed with a RBF (DX5)-specific monoclonal antibody on lysates from untreated, Warts-depleted, or RBF-depleted S2R⁺ cells. IgG was used as a nonspecific control. Following immunoprecipitation, Western blots were probed with antibodies specific to dE2F2, Warts, RBF (DX2), and β-Tubulin. Western blot analysis reveals that dE2F2 is immunoprecipitated through RBF in untreated or Warts-depleted cells, but not in cells depleted of RBF.