Modifiers of notch transcriptional activity identified by genome-wide RNAi
© Mourikis et al; licensee BioMed Central Ltd. 2010
Received: 30 March 2010
Accepted: 19 October 2010
Published: 19 October 2010
The Notch signaling pathway regulates a diverse array of developmental processes, and aberrant Notch signaling can lead to diseases, including cancer. To obtain a more comprehensive understanding of the genetic network that integrates into Notch signaling, we performed a genome-wide RNAi screen in Drosophila cell culture to identify genes that modify Notch-dependent transcription.
Employing complementary data analyses, we found 399 putative modifiers: 189 promoting and 210 antagonizing Notch activated transcription. These modifiers included several known Notch interactors, validating the robustness of the assay. Many novel modifiers were also identified, covering a range of cellular localizations from the extracellular matrix to the nucleus, as well as a large number of proteins with unknown function. Chromatin-modifying proteins represent a major class of genes identified, including histone deacetylase and demethylase complex components and other chromatin modifying, remodeling and replacement factors. A protein-protein interaction map of the Notch-dependent transcription modifiers revealed that a large number of the identified proteins interact physically with these core chromatin components.
The genome-wide RNAi screen identified many genes that can modulate Notch transcriptional output. A protein interaction map of the identified genes highlighted a network of chromatin-modifying enzymes and remodelers that regulate Notch transcription. Our results open new avenues to explore the mechanisms of Notch signal regulation and the integration of this pathway into diverse cellular processes.
The Notch (N) cell-surface receptor is the central element of one of the handful of signaling pathways that are evolutionary conserved throughout metazoans. Notch signaling directs the development of multicellular organisms through membrane-anchored interactions between adjacent cells. The response to Notch signals varies greatly and can result in diverse biological consequences, such as cell proliferation, differentiation or apoptosis. Notch signaling is initiated by the binding of the transmembrane Notch receptor with one of its ligands, Delta or Serrate, expressed on the surface of a neighboring cell . The receptor-ligand interaction induces a series of proteolytic events that releases the Notch intracellular domain (Nicd), which then translocates to the nucleus and complexes with transcription factors and co-activators to regulate target gene expression. Nicd, together with Suppressor of Hairless [Su(H)], a DNA binding protein in the CSL (CBF1/Su(H)/Lag2) family, and mastermind (mam) proteins form the core transcriptional complex of the signaling pathway, with the Enhancer of Split [E(spl)] locus being the most thoroughly characterized downstream transcriptional target. The Notch signaling pathway is modulated at many levels: Notch protein abundance, trafficking, and post-translational processing, as well as the regulated formation of repressive and promoting complexes on the DNA. The final cell fate outcome is a complex interplay between all the cellular factors that regulate Notch activity.
We designed a genome-wide RNA interference (RNAi) screen using a Drosophila cell culture-based system aimed to identify novel proteins that directly influence the signaling capacity of the core Notch pathway. This genome-wide RNAi screen was performed on Drosophila Kc167 cell cultures that were transfected with a construct that expresses a constitutively active, membrane-tethered form of the Notch receptor, NΔecn . Notch pathway activity was monitored by measuring the transcriptional response of a luciferase-reporter gene cassette (m3-luc) containing the native promoter element of the E(spl)m3 gene , the most Notch responsive E(spl) target in cell culture .
Cellular distribution of Notch modifiers selected by the complimentary analysis methods.
NΔecn > m3-luc/con-luc (Activators)
NΔecn > m3-luc/con-luc (Repressors)
NΔecn > m3-luc/m3-luc (Activators)
NΔecn > m3-luc/m3-luc (Repressors)
Total (discounting duplicates)
To further analyze and categorize our dataset, the Notch signaling modifiers identified in the study were combined with physical interaction data from public databases. The interaction map that was generated revealed classes of interacting Notch modifiers such as mRNA processing and ribosomal proteins. The network analysis also highlighted a central core of chromatin regulating genes, including chromatin modifying enzymes and remodelers that interact directly with the Su(H) DNA binding complex.
Results and Discussion
Development of a robust assay to measure changes in Notch transcriptional activity
A reporter assay was developed to measure Notch activity in a high-throughput Drosophila cell-based approach. The assay consists of three components: 1) a Notch activity reporter construct with two, tandem copies of the E(spl)m3 promoter positioned upstream of the firefly luciferase gene (m3-luc) ; 2) the constitutively active, membrane-tethered form of the Notch receptor with the extracellular domain removed (NΔecn), driven by the viral OpIE2 promoter; 3) a control construct that constitutively expresses firefly luciferase, also driven by the viral OpIE2 promoter (con-luc). Con-luc was used to normalize signal intensity relative to transfection efficiency, cell density and viability, and general effects on OpIE2-mediated transcription.
As predicted, we found that targeting Su(H) and mam with RNAi in cells expressing activated Notch resulted in a sharp reduction of the reporter luciferase activity (Figure 1A and 1B). Conversely, knock-down of Su(H) increased the basal activity of the m3-luc reporter in the absence of NΔecn (Figure 1C). These opposing effects of Su(H) RNAi on E(spl)m3 expression are consistent with the dual roles of Su(H) as a transcriptional repressor in the absence of Notch activation, as well as a transcriptional activator when complexed with processed Nicd in the nucleus . In contrast, RNAi against Hairless resulted in a marked decrease in the ratio of induced:uninduced signal of m3-luc (Figure 1B). This decrease is expected due to the specific de-repression of the uninduced Notch target promoter when H is knocked down (Figure 1C) and shows that there is robust Su(H)/H complex repressor activity in the uninduced Kc167 cells. The different ratios for H RNAi treatment obtained by the two different normalization methods (Figures 1A and 1B) highlights the additional mechanistic information that can be deduced when normalizing by the uninduced E(spl)m3 promoter activity. Hairless acts as a repressor in the uninduced cells, but has no apparent role in Notch activated cells.
Splitting the cells into three different assays also allows the uninduced Notch target promoter measurement to be used as an alternative and specific control for Notch induced activity. This additional control flags dsRNA treatments that may specifically affect transcription of the viral OpIE2 promoter. RNAi treatment may modulate either the signal of interest and/or the control signal and the resulting ratios may be altered indistinguishably between these possibilities. Whereas this second control will sort a subset of these dsRNAs as definitively altering Notch target transcription (positive by both normalization methods).
The Notch activity assay responded in a predictable and specific manner to RNAi of known Notch signaling components, and these data establish our experimental set-up (robotic cell transfer, normalization methods, incubation time) as optimal for detecting changes in Notch transcriptional activities.
Genome-wide RNAi screen and data analysis
The RNAi screen was performed using a dsRNA library from the Drosophila RNAi Screening Center (DRSC), containing a total of 23,560 dsRNAs, targeting known and predicted gene products. After four days of RNAi treatment, cells were uniformly dispensed by robotic liquid handling into microplates containing the different transfection mixes (con-luc, m3-luc and NΔecn > m3-luc). Each assay was performed in duplicate, and firefly luciferase activity was measured 24 h after transfection.
Classification of the identified proteins
Classification of modifiers identified in the screen was based upon gene ontologies (GO terms) as reported by Flybase . These classes are shown as a percentage of genes with that GO term and median z-scores of that class (Additional file 3). Certain classes showed particularly significant z-scores. For instance, activators of Notch induced transcription as normalized by the control reporter (con-luc) contained 10 chromatin-associated factors, 6.5% of the hits, and 16 transcription factors, representing 10.5%. Both these classes have a median z-score of -2.9, placing these groups in the top 0.2% of the calculated genome-wide distribution (Additional file 3C). Of the identified genes, 90 have predicted and known human orthologs associated with human genetic disorders (Additional file 4) .
Known Notch pathway interactors found by the RNAi screening method
Thirteen genes that have been described to genetically interact with Notch were identified (Figure 2C, boxed gene names). Among these, the core Notch pathway transcription factor Su(H) and the repressor Hairless further validated the screening method. We also recovered the known negative regulator of Notch signaling, Suppressor of deltex [Su(dx)], encoding a cytoplasmic protein that functions as an E3 ubiquitin ligase that ubiquitinates membrane-anchored Notch , and prickle (pk), encoding a transcription factor known to play a role in E(spl)mδ gene expression . Nine other genes (heph, al, Sos, toe, Vg, Lobe, lid, Nipped-A and RpL19) were identified that have been shown to genetically interact with Notch signaling, but whose mechanistic level of integration into the Notch pathway are understood to varying degrees [9–17].
An in vivo RNAi screen for Notch activity has recently been published that is based on bristle and wing morphology and as a different approach to this transcriptional based study, the overlap was minimal . Of the 14 genes listed in the previous study that have known genetic interactions with Notch, only tramtrack (ttk) is common to both screens. The direct transcription based method of our study would be expected to be better suited to identify transcription and chromatin factors, as indicated by the strong scores of repressor components and core chromatin components identified (Additional file 3). In contrast, the phenotype-based study was more sensitive to membrane trafficking machinery, making the two studies complimentary.
Protein interaction network of Notch transcription modifiers
The importance of chromatin in Notch regulation has recently become apparent and this transcription-based screen was suited to uncover this class of regulators. On average, chromatin-modifying genes scored relatively high in the data analysis (Additional file 3). The interaction map reveals a central core of chromatin modifying components that have multiple physical connections to the nuclear elements of the Notch pathway such as Su(H) and H (Figure 3B). Many of these chromatin components are known to interact genetically and physically with the Notch pathway [14, 21–23].
The protein interaction network also shows a number of protein classes that have no known mechanistic link to Notch transcriptional regulation. For these classes of molecules (mRNA processing proteins and ribosomal factors, discussed later), the network suggests that they may be affecting Notch signaling through direct interactions with these core chromatin components (Figure 3D and 3E).
Epistatic analysis of candidate genes
Factors involved in chromatin modification
The transcription-based screening method using an endogenous E(spl)m3 promoter sequence was particularly useful for identifying chromatin components. We identified several chromatin factors previously shown to affect Notch-dependent transcription. A component of the SAGA histone acetyltransferase complex, Nipped-A, was identified. Nipped-A, the Drosophila homologue of yeast Tra1 and mammalian TRAP proteins, is a key factor of the SAGA complex. It has been shown previously that reduced Nipped-A dosage enhances the wing notching phenotype of both mastermind and Notch mutants [14, 25]. The RNAi treated cell culture data demonstrates that Nipped-A promotes transcription at the E(spl)m3 promoter both in the presence and absence of activated Notch (Figures 4 and 5). This shows that the result of Nipped-A function is independent of whether active Nicd is localized on the target promoter.
We also identified several homologues of components of the Rpd3 histone deacetylase co-repressor complex, including Sin3a, Sds3 (CG14220), a putative ortholog of SAP130 (Sin3A Associated Protein 130, CG11006), and Rpd3 itself (Figures 4, 5 and Additional file 1B). When these factors were targeted by RNAi, there was an increase in Notch-induced reporter transcription, consistent with the role of the Rpd3 complex and histone deacetylation as a transcriptional inhibitor . Conversely, knocking down Sin3a had the opposite effect on the uninduced baseline activity of the E(spl)m3 promoter (Figure 5C). Thus, unlike the histone acetylation complex (SAGA), the activity of the deacetylation complex (Sin3A) on the E(spl)m3 promoter is dependent on the presence of activated Notch.
The screen identified several components of the chromatin remodeling complex Brahma: Brm Associated Protein 55 (Bap55), Brm Associated Protein 170 (Bap170), polybromo, and moira (mor). A previous Drosophila phenotype based screen has found a genetic interaction between the Notch ligand Delta and another component of the Brahma complex, brahma (brm) . Loss of function brm alleles were found to enhance Delta mutant phenotypes in eye and bristle development . The various Brahma components identified in this study show a complex array of effects on the transcription of the E(spl)m3 promoter, some consistent with previously described loss of function brm alleles while others opposing. RNAi directed against Bap55 and polybromo demonstrated a specific reduction in Notch induced transcription (Figure 2C, area b) that is consistent with the previously observed role of brm in Notch signaling during Drosophila development . Unexpected are the Brahma subunits identified that modulate transcription from the uninduced E(spl)m3 promoter: Bap170 and mor. The screen reveals that both of these components specifically mediate transcription from the uninduced E(spl)m3 promoter, while Bap170 activates and mor represses (Figure 2C, area c.).
In addition to chromatin modifying complexes, a new interaction between the histone variant H3.3 and Notch signaling is seen. RNAi treatment of either genomic copy of the H3.3 histone variant (H3.3A and H3.3B) shows a dramatic decrease in Notch activated transcription (Figures 4 and 5). The histone variant H3.3 has been shown to be incorporated into the promoters of actively transcribed genes in a replication independent process to maintain transcription and its influence on Notch targeted transcription remains to be explored .
A major question that arises from these data is, how specific can the identified chromatin factors be to regulating Notch transcription? It has recently been noted that chromatin components are more selective in function than was previously thought. Surprisingly, there are now a handful of examples where modulating the expression of a single target gene can rescue the phenotype associated with a null mutation in a chromatin remodeling complex component . By immunoprecipitation and mass-spec analysis, it has recently been shown that the Notch repressor complex contains a host of chromatin modifying components . These identified components include Sin3A, Rpd3, lid, Bap55 and moira, factors that were also uncovered in this screen as modifiers of Notch target transcription. This repressor complex has been shown to be recruited to Notch target promoters by Su(H) and this interaction may provide a mechanism for targeting the activity of these chromatin factors to Notch signaling [22, 23]. This is consistent with the observation that the genetic interactions demonstrated between this repressor complex and Notch were not seen when tested against a host of other signaling pathways [22, 23]. Control reporter transcription levels in this study indicated that targeting these chromatin genes by dsRNA did not significantly reduce cell viability and growth over the course of the five-day RNAi incubation in culture. The screen data shows that Notch signaling may be particularly sensitive to the levels of these chromatin components in the cell, while the protein interaction network confirms that many of these chromatin factors physically interact with Su(H) and Hairless suggesting a mechanism to explain this observation.
Regulation of histone position and modification are known factors that determine the "context dependent" nature of Notch signaling during development. These factors differentially interpret the signals received from the cell surface by recording an epigenetic history on the target promoter. This transcription-based screen revealed new chromatin factors that can be further studied for their role in Notch-mediated development.
mRNA processing factors
The genome-wide transcription assay revealed two other classes of proteins not conventionally associated with transcriptional regulation. A number of ribosomal components and proteins associated with mRNA processing were found to regulate transcription of the activated Notch target gene (Figures 2 and 3). What is unexpected about these interactions is their relative specificity, as was for the chromatin components. Again, any RNAi treatments in the genome screen that significantly effected cell viability or general transcription were excluded from the analysis. In addition, all Notch induced target transcription signals were subsequently normalized to either the control signal or the uninduced Notch target promoter. This analysis demonstrated that knocking down these components of the ribosome and splicing machinery did not significantly affect general cell viability and had a relatively specific effect on Notch target transcription.
A number of mRNA splicing and processing components were found to interact with Notch-activated transcription (Figure 2A and 2C). As expected, these proteins demonstrated extensive physical interactions with each other (Figure 3D). Unexpectedly, these mRNA modifying proteins show physical interactions with the core chromatin components identified in this transcription based screen (Figure 3D). The polypyrimidine tract binding proteins Sex lethal (Sxl) and hephaestus (heph) were found to repress and activate Notch promoter activity, respectively, in our cell culture assay. Heph was previously found to interact genetically with Notch signaling during wing development . Other mRNA processing components, such as the non-sense mediated decay factors Upf1, Upf2 and Smg1, were found to modulate Notch activated transcription in the analysis (Figures 2 and 3D). These mRNA components may be interacting indirectly with Notch transcription through their mRNA processing functions - for instance, by specifically controlling the mRNA processing of transcripts for an essential Notch signaling factor such as Su(H). The network suggests a possible alternate mechanism to explain the interaction between the identified mRNA processing factors and Notch transcription, one that is mediated though the chromatin machinery (Figure 3D).
In plants, components of the nuclear cap-binding complex (including Cbp20) functionally interact with microRNA (miRNA) processing components, such as Ars2, giving these proteins dual roles in splicing and miRNA processing . The role of Cbp20 in miRNA processing was also confirmed in Drosophila and mammalian systems [30, 31]. The nuclear cap-binding complex component Cbp20 was found to mediate Notch transcription in this study (Figure 2) and demonstrates physical interactions with the chromatin remodeling component Ssrp (Figure 3D). The interaction network suggests that the miRNA processing activity of Cbp20 may be targeted to Notch signaling through interactions with the chromatin remodeling machinery.
Ribosomal factors and the classical Minutemutations
A complex of ribosomal proteins was identified that modulated Notch reporter transcription (Figures 2 and 3E). This class of translation factors included the large ribosomal subunit RpL19 that belongs to the Minute genetic class. The Minutes are a class of ribosomal gene mutations that are homozygous lethal, delay cellular growth when heterozygous and have a rich history of study [32, 33]. Of interest, RpL19 has been shown to be a modifier of Notch signaling [15, 34, 35]. In fact, the Minute class of genes was first described in detail as modifiers of Notch signaling in 1929 . Since then, ribosomal components have been widely observed as effectors of Notch [25, 37–39]. The Notch transcription reporter measurements compliment these long-standing, yet mechanistically unknown, genetic interactions. One mechanism proposed to explain the relatively specific genetic interactions between Minute mutations and Notch, is the possibility of specific translational effects. For instance, the translation of long transcripts such as the one encoding Notch itself may be sensitive to lower levels of specific ribosomal components. In contrast, an alternative hypothesis has been presented that these ribosomal proteins may have post-translational effects on key components of Notch signaling . Minute protein mutations are not found in the active site of the ribosome, as the peptide synthesis reaction is catalyzed exclusively by RNA in the core, but rather on the surface of the ribosome. Current structural and biochemical studies have demonstrated post-translational roles for these surface coating ribosomal proteins . This includes the folding of nascent peptide chains either directly on the surface of the ribosome or by the co-recruitment of protein chaperones. The protein-protein interaction map suggests that these types of post-translational interactions may be directed towards the core chromatin components of the Notch network (Figure 5E). Such a direct mechanism could explain the transcriptional effects described in this study, as well as the long-standing genetic observations between Notch and the Minute class of mutations.
Transcription factors that affect Notch-dependent transcription
Analysis of the genes identified in the screen revealed a number of transcription factors that affect Notch-dependent transcription. Among these are cnc and maf-S that are known to form a strong transcriptional activator complex . RNAi targeting of either of these two genes strongly suppressed both the Notch-induced as well as non-induced E(spl)m3 reporter activity (Additional file 1A). Also, among the 15 transcription factors that promote Notch activity, we found the DNA binding protein Deaf-1. Cnc, maf-S, and Deaf-1 are reported to interact with the Hox protein Deformed (Dfd) to regulate segmentation, but their roles in other developmental events are not known . Our results provide a possible role of these proteins in Drosophila development by promoting Notch signaling.
Another transcription factor that we found to play an agonistic role in Notch signaling is the homeobox containing protein Aristaless (al) (Figure 5B). Al has been tentatively linked to Notch signaling, as it cell autonomously represses the Notch ligand Delta in the pretarsus during leg morphogenesis . It is possible that al is involved in a Notch-mediated lateral inhibition mechanism, where al expressing cells remain undifferentiated by favoring active Notch signaling whereas their neighboring cells are free to express Delta and differentiate. It has also been shown that Notch mutant clones in the developing leg disk show diminished al levels, suggesting that al is a Notch target gene. This would be the predicted relationship in a lateral inhibition system, where a Notch/al positive feedback loop would amplify the Notch activity differences between neighboring cells.
Two additional transcription factors that have been previously shown to be involved in leg morphogenesis were found to promote Notch signaling: Bonus (bon), a homologue of the vertebrate TIF1beta transcriptional cofactor , and crooked legs (croI), a zinc finger protein . Notch signaling is known to play an important role in Drosophila leg development, and the recovery of these two transcription factors as modifiers of Notch-induced E(spl)m3 expression suggests that bon and croI may function to modulate Notch target gene output in the developing leg .
We also identified the Drosophila orthologues of two mammalian proto-oncogenes kayak (c-fos), and c-Myb, as positive regulators of Notch-signaling. Although a direct functional link between these proteins and Notch signaling has not been described, kayak has been shown to interact genetically with Hairless  and c-Myb genetically interacts with bon, a novel Notch modifier described above . In addition, our data reveals a synergistic relationship between the positive regulator of Ras signaling, 14-3-3ε, and Notch. Once again, the protein interaction network shows extensive contacts between 14-3-3ε and the chromatin machinery, suggesting a mechanism for modulating Notch target transcription through Su(H) mediated chromatin modifications. Interactions between Notch and oncogenic pathways are of particular interest, as the involvement of Notch in cancer biology and stem cell maintenance is becoming increasingly apparent.
An unexpected Notch target transcription modifier identified in the screen is the Notch target gene Tramtrack (ttk). We found that targeting of ttk with dsRNA resulted in reduced Notch activity (Figures 4 and 5). In contrast, ttk expression itself has been shown to increase in response to ectopic Notch activity . The RNAi treatment data suggest that ttk may function in a positive feedback mechanism to promote Notch activated transcription and the network analysis suggests that this interaction may be mediated by a direct contact with Notch itself (Figure 3C).
A complementary, genome-wide RNAi approach has revealed a subset of factors that modulate Notch target transcription that may not have been found by traditional genetic approaches. For instance, pleiotropic effects combined with non-saturating mutagenesis may have obscured the detection of some components in traditional genetic screens. Several novel modifiers were identified in this RNAi transcription-based screen, and these factors will be further investigated for their precise roles in the regulation of Notch signaling during development. In addition, the interaction network of these factors suggests that many may work through contacts with chromatin machinery components that are in turn directed to Notch target promoters by the transcription factor Su(H).
Constitutively active Notch constructs were made with cDNA encoding either membrane-tethered, Drosophila Notch (NΔecn), constitutively activated by the removal of the extracellular domain, or the soluble intracellular domain (Nicd) . These truncated Notch constructs were cloned into the pIZ-V5/His expression vector (Invitrogen) producing non-tagged proteins. The control expression plasmid (con-luc) was constructed by cloning firefly luciferase into the same pIZ-V5/His expression vector. The luciferase reporter construct (m3-luc) contains a 1.4 kb tandem duplication of E(spl)m3 upstream regulatory sequences, cloned into a pGL2-Basic vector (Promega), as described .
Genome-wide RNAi method
A total of 23,560 dsRNAs, made available from the Drosophila RNAi Screening Center (DRSC) at Harvard Medical School, were screened by the following method: Kc167 cells were washed three times and resuspended in serum-free Sang's M3 medium (Sigma) at a concentration of 5 × 105 cells/ml. Using a robotic liquid handler, 104 cells (20 ml) were uniformly dispensed into the wells of 384-well polypropylene plates containing dsRNA and incubated for 45 min at room temperature. An equal volume of M3 medium containing 10% fetal bovine serum was added and incubated for four days. On day four, the RNAi-treated cells were diluted with 100 μl of medium, mixed and 20 μl were dispensed into the wells of six new 384-well plates, pre-aliquoted with 20 μl of transfection mix. The six plates contained the three different transfection mixes, each in duplicate. Transfection mixes were prepared with Effectene Transfection Reagent (Qiagen), following the manufacturer's guidelines. Luciferase activity was measured 24 h post transfection using the Steady-Glo Luciferase Assay System (Promega).
This method requires only two plasmids to be transfected at one time and gave acceptable signal to noise ratios for high throughput screening in 384-well plate format (Figure 1). Whereas, the conventional renilla dual-glo assay was not robust enough to scale to 384 well format with this Notch reporter system using an endogenous target. In contrast to pathways with soluble ligands, the reporter and constitutively active Notch constructs are required to transfect the same cell to activate transcription. With the renilla dual-glo system, adding the control construct required the co-transfection of three individual plasmids and this reduced the signal to noise ratio to insufficient levels.
Duplicate measurements for each of the three signals were averaged; Notch specific E(spl)m3 promoter in the presence of activated Notch (NΔecn > m3-luc), the E(spl)m3 promoter alone (m3-luc) and the unrelated viral promoter OplE2 (con-luc). The NΔecn > m3-luc signal was normalized two different ways, by either the m3-luc or con-luc signals. The z-scores of the log2 ratios were calculated by using the standard deviation and mean of the measurements that corresponded to the 96-wells of the original dsRNA stock plates . To remove ratios that contained data that did not sufficiently replicate in the original duplicate measurements, a distribution was calculated for the individual errors (estimated from the duplicate data sets) and ratios with an associated error z-score above 3 were removed from further analysis (0.9% of the total data).
To remove data associated with dsRNA that greatly reduced general transcription or cell viability, a distribution of the signals from the control promoter (con-luc) was calculated, and data with z-scores below -2 were removed (4% of the total wells). All calculations were done by in house software written in JAVA.
Hits were chosen as those log2 ratios with a z-score above 2 or below -2 for NΔecn > m3-luc normalized by the viral promoter OplE2 (con-luc). For the data set normalized by the E(spl)m3 promoter alone (m3-luc), a z-score above 1.8 or below -1.8 was used. The m3-luc normalized distribution had more defined outliers indicating a better data set. As a consequence, m3-luc normalized data distribution had higher kurtosis as seen by a slightly sharper peak in Figure 2. This does not change the rank order or relative differences in the hits of that data set, but to make the cut-offs more equivalent between the two normalization methods, the different cut-off values were used.
RNAi retest procedure
Genes were chosen for retesting that were selected as positive by both normalization methods (Figure 2c, area a.). This second set of 28 dsRNAs were independently redesigned by the method of Arziman et al. with no predicted off targets and are listed in Additional file 5.
DNA templates for T7 reactions were generated by PCR from Kc167 cell genomic DNA and dsRNA was produced using the MEGAscript RNAi kit (Ambion). Per well, 25 ml of Kc167 cells at a concentration of 8 × 105 cells/ml were incubated with 1.25 μg of dsRNA (0.5 mg/ml stock) for 1 h in serum-free M3 medium. M3 medium with 10% FBS (75 μl) was then added and incubated for 4 days. On the fourth day, 125 μl of medium was added, and treated cells were split into 4 wells with 50 μl per well, each containing 50 μl of the following transfection mixes, prepared as above: a. con-luc, b. m3-luc, c. m3-luc & pIZ-NΔecn, d. m3-luc & pIZ-Nicd. Luciferase levels were measured after 25 h, as above. Retests were done in quadruplicate for each dsRNA, and the results are given in Additional file 5 for the 22 positive retests that have p-values < 0.05 (compared with control dsRNA treated cells).
Notch interaction network construction
The Notch interaction network was generated by combining physical interaction data (e.g. two-hybrid data) from the DroID database  (that contained interactions from the BioGRID database ) with Notch transcription modifiers identified in the genome-wide study. Genetic interactions were not used for the network map. The resulting network was drawn using Cytoscape  and the data can be found in additional file 6. The network file can be viewed in detail using the open source Cytoscape viewer http://www.cytoscape.org.
All orthology predictions for candidate genes were made using InParanoid .
We thank S. Artavanis-Tsakonas for generous support of this work. For providing the RNAi library, we thank N. Perrimon and the Drosophila RNAi Screening Center (DRSC). Screening resources and expertise were provided by C. Shamu and the ICCB - Longwood screening facility.
This work was supported by NIH grants HG003616 and NS26084 under S. Artavanis-Tsakonas. The Drosophila RNAi Screening Center (DRSC) is supported by grant 2R01GM067761-05 from the NIGMS division of NIH.
- Artavanis-Tsakonas S, Rand M, Lake R: Notch signaling: cell fate control and signal integration in development. Science. 1999, 284: 770-6. 10.1126/science.284.5415.770.View ArticlePubMedGoogle Scholar
- Rebay I, Fehon RG, Artavanis-Tsakonas S: Specific truncations of Drosophila Notch define dominant activated and dominant negative forms of the receptor. Cell. 1993, 74: 319-29. 10.1016/0092-8674(93)90423-N.View ArticlePubMedGoogle Scholar
- Mukherjee A, Veraksa A, Bauer A, Rosse C, Camonis J, Artavanis-Tsakonas S: Regulation of Notch signalling by non-visual beta-arrestin. Nat Cell Biol. 2005, 7: 1191-201. 10.1038/ncb1327.View ArticlePubMedGoogle Scholar
- Krejcí A, Bray S: Notch activation stimulates transient and selective binding of Su(H)/CSL to target enhancers. Genes & Development. 2007, 21: 1322-7. 10.1101/gad.424607.View ArticleGoogle Scholar
- Wilson R, Goodman J, Strelets V: FlyBase: integration and improvements to query tools. Nucleic Acids Res. 2008, 36: D588-93. 10.1093/nar/gkm930.PubMed CentralView ArticlePubMedGoogle Scholar
- Reiter L, Potocki L, Chien S, Gribskov M, Bier E: A systematic analysis of human disease-associated gene sequences in Drosophila melanogaster. Genome Res. 2001, 11: 1114-25. 10.1101/gr.169101.PubMed CentralView ArticlePubMedGoogle Scholar
- Fostier M, Evans D, Artavanis-Tsakonas S, Baron M: Genetic characterization of the Drosophila melanogaster Suppressor of deltex gene: A regulator of notch signaling. Genetics. 1998, 150: 1477-85.PubMed CentralPubMedGoogle Scholar
- Cooper M, Bray S: Frizzled regulation of Notch signalling polarizes cell fate in the Drosophila eye. Nature. 1999, 397: 526-30. 10.1038/17395.View ArticlePubMedGoogle Scholar
- Abu-Issa R, Cavicchi S: Genetic interactions among vestigial, hairy, and Notch suggest a role of vestigial in the differentiation of epidermal and neural cells of the wing and halter of Drosophila melanogaster. J Neurogenet. 1996, 10: 239-46. 10.3109/01677069609083465.View ArticlePubMedGoogle Scholar
- Campbell G: Regulation of gene expression in the distal region of the Drosophila leg by the Hox11 homolog, C15. Dev Biol. 2005, 278: 607-18. 10.1016/j.ydbio.2004.12.009.View ArticlePubMedGoogle Scholar
- Chern J, Choi K: Lobe mediates Notch signaling to control domain-specific growth in the Drosophila eye disc. Development. 2002, 129: 4005-13.PubMedGoogle Scholar
- Dansereau D, Lunke M, Finkielsztein A, Russell M, Brook W: Hephaestus encodes a polypyrimidine tract binding protein that regulates Notch signalling during wing development in Drosophila melanogaster. Development. 2002, 129: 5553-66. 10.1242/dev.00153.View ArticlePubMedGoogle Scholar
- Dominguez M, Ferres-Marco D, Gutierrez-Avino F, Speicher S, Beneyto M: Growth and specification of the eye are controlled independently by Eyegone and Eyeless in Drosophila melanogaster. Nat Genet. 2004, 36: 31-9. 10.1038/ng1281.View ArticlePubMedGoogle Scholar
- Gause M, Eissenberg JC, Macrae AF, Dorsett M, Misulovin Z, Dorsett D: Nipped-A, the Tra1/TRRAP subunit of the Drosophila SAGA and Tip60 complexes, has multiple roles in Notch signaling during wing development. Mol Cell Biol. 2006, 26: 2347-59. 10.1128/MCB.26.6.2347-2359.2006.PubMed CentralView ArticlePubMedGoogle Scholar
- Hart K, Klein T, Wilcox M: A Minute encoding a ribosomal protein enhances wing morphogenesis mutants. Mech Dev. 1993, 43: 101-10. 10.1016/0925-4773(93)90028-V.View ArticlePubMedGoogle Scholar
- Verheyen E, Purcell K, Fortini M, Artavanis-Tsakonas S: Analysis of dominant enhancers and suppressors of activated Notch in Drosophila. Genetics. 1996, 144: 1127-41.PubMed CentralPubMedGoogle Scholar
- Liefke R, Oswald F, Alvarado C, Ferres-Marco D, Mittler G, Rodriguez P, Dominguez M, Borggrefe T: Histone demethylase KDM5A is an integral part of the core Notch-RBP-J repressor complex. Genes & Development. 2010, 24: 590-601. 10.1101/gad.563210.View ArticleGoogle Scholar
- Mummery-Widmer JL, Yamazaki M, Stoeger T, Novatchkova M, Bhalerao S, Chen D, Dietzl G, Dickson BJ, Knoblich JA: Genome-wide analysis of Notch signalling in Drosophila by transgenic RNAi. Nature. 2009, 458: 987-92. 10.1038/nature07936.PubMed CentralView ArticlePubMedGoogle Scholar
- Yu J, Pacifico S, Liu G, Finley RL: DroID: the Drosophila Interactions Database, a comprehensive resource for annotated gene and protein interactions. BMC Genomics. 2008, 9: 461-10.1186/1471-2164-9-461.PubMed CentralView ArticlePubMedGoogle Scholar
- Breitkreutz BJ, Stark C, Reguly T, Boucher L, Breitkreutz A, Livstone M, Oughtred R, Lackner DH, Bähler J, Wood V, et al: The BioGRID Interaction Database: 2008 update. Nucleic Acids Res. 2008, 36: D637-40. 10.1093/nar/gkm1001.PubMed CentralView ArticlePubMedGoogle Scholar
- Armstrong JA, Sperling AS, Deuring R, Manning L, Moseley SL, Papoulas O, Piatek CI, Doe CQ, Tamkun JW: Genetic screens for enhancers of brahma reveal functional interactions between the BRM chromatin-remodeling complex and the delta-notch signal transduction pathway in Drosophila. Genetics. 2005, 170: 1761-74. 10.1534/genetics.105.041327.PubMed CentralView ArticlePubMedGoogle Scholar
- Moshkin YM, Kan TW, Goodfellow H, Bezstarosti K, Maeda RK, Pilyugin M, Karch F, Bray SJ, Demmers JAA, Verrijzer CP: Histone chaperones ASF1 and NAP1 differentially modulate removal of active histone marks by LID-RPD3 complexes during NOTCH silencing. Mol Cell. 2009, 35: 782-93. 10.1016/j.molcel.2009.07.020.View ArticlePubMedGoogle Scholar
- Goodfellow H, Krejcí A, Moshkin Y, Verrijzer CP, Karch F, Bray SJ: Gene-specific targeting of the histone chaperone asf1 to mediate silencing. Developmental Cell. 2007, 13: 593-600. 10.1016/j.devcel.2007.08.021.View ArticlePubMedGoogle Scholar
- Herranz H, Stamataki E, Feiguin F, Milan M: Self-refinement of Notch activity through the transmembrane protein Crumbs: modulation of gamma-secretase activity. EMBO Rep. 2006, 7: 297-302. 10.1038/sj.embor.7400617.PubMed CentralView ArticlePubMedGoogle Scholar
- Rollins R, Morcillo P, Dorsett D: Nipped-B, a Drosophila homologue of chromosomal adherins, participates in activation by remote enhancers in the cut and Ultrabithorax genes. Genetics. 1999, 152: 577-93.PubMed CentralPubMedGoogle Scholar
- Wolffe A: Histone deacetylase: a regulator of transcription. Science. 1996, 272: 371-2. 10.1126/science.272.5260.371.View ArticlePubMedGoogle Scholar
- Ahmad K, Henikoff S: The histone variant H3.3 marks active chromatin by replication-independent nucleosome assembly. Mol Cell. 2002, 9: 1191-200. 10.1016/S1097-2765(02)00542-7.View ArticlePubMedGoogle Scholar
- Ho L, Crabtree GR: Chromatin remodelling during development. Nature. 2010, 463: 474-84. 10.1038/nature08911.PubMed CentralView ArticlePubMedGoogle Scholar
- Laubinger S, Sachsenberg T, Zeller G, Busch W, Lohmann JU, Rätsch G, Weigel D: Dual roles of the nuclear cap-binding complex and SERRATE in pre-mRNA splicing and microRNA processing in Arabidopsis thaliana. Proc Natl Acad Sci USA. 2008, 105: 8795-800. 10.1073/pnas.0802493105.PubMed CentralView ArticlePubMedGoogle Scholar
- Gruber JJ, Zatechka DS, Sabin LR, Yong J, Lum JJ, Kong M, Zong WX, Zhang Z, Lau C-K, Rawlings J, et al: Ars2 links the nuclear cap-binding complex to RNA interference and cell proliferation. Cell. 2009, 138: 328-39. 10.1016/j.cell.2009.04.046.PubMed CentralView ArticlePubMedGoogle Scholar
- Sabin LR, Zhou R, Gruber JJ, Lukinova N, Bambina S, Berman A, Lau C-K, Thompson CB, Cherry S: Ars2 regulates both miRNA- and siRNA- dependent silencing and suppresses RNA virus infection in Drosophila. Cell. 2009, 138: 340-51. 10.1016/j.cell.2009.04.045.PubMed CentralView ArticlePubMedGoogle Scholar
- Morata G, Ripoll P: Minutes: mutants of drosophila autonomously affecting cell division rate. Dev Biol. 1975, 42: 211-21. 10.1016/0012-1606(75)90330-9.View ArticlePubMedGoogle Scholar
- Lambertsson A: The minute genes in Drosophila and their molecular functions. Adv Genet. 1998, 38: 69-134. full_text.View ArticlePubMedGoogle Scholar
- Hall LE, Alexander SJ, Chang M, Woodling NS, Yedvobnick B: An EP overexpression screen for genetic modifiers of Notch pathway function in Drosophila melanogaster. Genet Res. 2004, 83: 71-82. 10.1017/S0016672304006731.View ArticlePubMedGoogle Scholar
- Klein T, Campos-Ortega JA: Second-site modifiers of the Delta wing phenotype in Drosophila melanogaster. Roux's Archives of Developmental Biology. 1992, 202: 49-60. 10.1007/BF00364596.View ArticleGoogle Scholar
- Schultz J: The Minute Reaction in the Development of DROSOPHILA MELANOGASTER. Genetics. 1929, 14: 366-419.PubMed CentralPubMedGoogle Scholar
- Dietrich U, Campos-Ortega JA: The expression of neurogenic loci in imaginal epidermal cells of Drosophila melanogaster. J Neurogenet. 1984, 1: 315-32. 10.3109/01677068409107094.View ArticlePubMedGoogle Scholar
- Röttgen G, Wagner T, Hinz U: A genetic screen for elements of the network that regulates neurogenesis in Drosophila. Mol Gen Genet. 1998, 257: 442-51. 10.1007/s004380050668.View ArticlePubMedGoogle Scholar
- Schmidt A, Hollmann M, Schäfer U: A newly identified Minute locus, M(2)32D, encodes the ribosomal protein L9 in Drosophila melanogaster. Mol Gen Genet. 1996, 251: 381-7.PubMedGoogle Scholar
- Cabrita LD, Dobson CM, Christodoulou J: Protein folding on the ribosome. Current Opinion in Structural Biology. 2010, 20: 33-45. 10.1016/j.sbi.2010.01.005.View ArticlePubMedGoogle Scholar
- Veraksa A, McGinnis N, Li X, Mohler J, McGinnis W: Cap 'n' collar B cooperates with a small Maf subunit to specify pharyngeal development and suppress deformed homeotic function in the Drosophila head. Development. 2000, 127: 4023-37.PubMedGoogle Scholar
- Veraksa A, Kennison J, McGinnis W: DEAF-1 function is essential for the early embryonic development of Drosophila. genesis. 2002, 33: 67-76. 10.1002/gene.10090.View ArticlePubMedGoogle Scholar
- Beckstead R, Ortiz J, Sanchez C, Prokopenko S, Chambon P, Losson R, Bellen H: Bonus, a Drosophila homolog of TIF1 proteins, interacts with nuclear receptors and can inhibit betaFTZ-F1-dependent transcription. Mol Cell. 2001, 7: 753-65. 10.1016/S1097-2765(01)00220-9.PubMed CentralView ArticlePubMedGoogle Scholar
- D'Avino P, Thummel C: crooked legs encodes a family of zinc finger proteins required for leg morphogenesis and ecdysone-regulated gene expression during Drosophila metamorphosis. Development. 1998, 125: 1733-45.PubMedGoogle Scholar
- de Celis J, Tyler D, de Celis J, Bray S: Notch signalling mediates segmentation of the Drosophila leg. Development. 1998, 125: 4617-26.PubMedGoogle Scholar
- Muller D, Kugler S, Preiss A, Maier D, Nagel A: Genetic modifier screens on Hairless gain-of-function phenotypes reveal genes involved in cell differentiation, cell growth and apoptosis in Drosophila melanogaster. Genetics. 2005, 171: 1137-52. 10.1534/genetics.105.044453.PubMed CentralView ArticlePubMedGoogle Scholar
- Nomura T, Tanikawa J, Akimaru H, Kanei-Ishii C, Ichikawa-Iwata E, Khan M, H Ito, Ishii S: Oncogenic activation of c-Myb correlates with a loss of negative regulation by TIF1beta and Ski. J Biol Chem. 2004, 279: 16715-26. 10.1074/jbc.M313069200.View ArticlePubMedGoogle Scholar
- Guo M, Jan L, Jan Y: Control of daughter cell fates during asymmetric division: interaction of Numb and Notch. Neuron. 1996, 17: 27-41. 10.1016/S0896-6273(00)80278-0.View ArticlePubMedGoogle Scholar
- Quackenbush J: Microarray data normalization and transformation. Nat Genet. 2002, 32 (Suppl): 496-501. 10.1038/ng1032.View ArticlePubMedGoogle Scholar
- Arziman Z, Horn T, Boutros M: E-RNAi: a web application to design optimized RNAi constructs. Nucleic Acids Res. 2005, 33: W582-8. 10.1093/nar/gki468.PubMed CentralView ArticlePubMedGoogle Scholar
- Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003, 13: 2498-504. 10.1101/gr.1239303.PubMed CentralView ArticlePubMedGoogle Scholar
- Berglund AC, Sjölund E, Ostlund G, Sonnhammer ELL: InParanoid 6: eukaryotic ortholog clusters with inparalogs. Nucleic Acids Res. 2008, 36: D263-6. 10.1093/nar/gkm1020.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.