Final Report Summary - CANCERPATHWAYS (Developmental Molecular Pathways in Drosophila as a Model for Human Cancer)
Cancer is one of the leading causes of death in Europe and worldwide. Although significant advances have been made in cancer research, it still remains a public health concern with a survival rate of about 50 %. It is therefore of great importance to develop new and specific drug treatments for cancer. The EU collaborative project CANCERPATHWAYS combines scientific expertise of eight partners to produce fundamental advances in our understanding of cancer biology and how this disease can be attacked.
Tumour development is characterised by the uncontrolled growth of cells. The deregulation of diverse signal transduction pathways has previously been linked with carcinogenesis. These cancer pathways include Wnt, Notch, Ras, Hippo, PI3K and JAK/STAT signalling cascades. Although new therapeutic approaches in cancer treatment target signalling pathway components, the full potential of signalling molecules as therapy targets remains to be explored. The CANCERPATHWAYS consortium aimed to identify novel targets and drug-like molecules for therapeutic application. In this respect we made use of the high evolutionary conservation of most oncogenic pathways with Drosophila representing a particularly powerful model system for the analysis of signalling cascades.
The CANCERPATHWAYS project integrated recent technological advances, such as genome-wide RNAi and compound screens in cell-based and in vivo models as well as computational approaches. We have established a comprehensive set of cell-based and in vivo bioassays for diverse oncogenic signalling pathways. Novel assays were developed for screening of small molecular compound libraries for their impact on cell fitness, specific pathway activities, tumour growth and metastasis in Drosophila. These assays were successfully validated and applied to perform high-throughput cell-based and in vivo RNAi and compound screens in Drosophila.
RNAi has proven to be a valid tool in Drosophila to systematically identify signalling pathway components. However, experiments with previous RNAi libraries had inherent false-positive and false-negative rates. Thus, we computationally designed and synthesised novel genome-wide RNAi libraries for cell-based and in vivo RNAi screening in Drosophila. Furthermore, high-throughput assays generate large datasets that need to be processed in order to identify specific hits for both, RNAi and compound screening experiments. One important point of the project was the strong integration of bioinformatics tools with the screening procedure. We have developed and improved data analysis tools and databases displaying the screening results. All these resources are of immense use and were made available for the research community.
Applying the technological advancements of the CANCERPATHWAYS project, we have identified several novel candidate genes and demonstrated their function in cancer signalling pathways. Many of those studies are already published by the CANCERPATHWAYS consortium while further studies will be published soon. High-throughput compound screens in Drosophila cells and in vivo tumour models resulted in the identification of known anti-cancer drugs as well as novel compounds that interfere with the activity of oncogenic signalling pathways in vitro and in vivo models.
Project context and objectives:
Cancer is one of the leading causes of death in Europe and worldwide. Although significant advances have been made in cancer research, it still remains a public health concern. It is therefore of great importance to develop new and specific drug treatments for cancer. The EU medium scale focused research project CANCERPATHWAYS combines scientific expertise of eight European partners to produce fundamental advances in our understanding of cancer biology and how this disease can be attacked.
Drosophila as a model system to study cancer
Cancer has often been characterised as a disease of signal transduction pathways. During development and homeostasis, cells constantly receive signals from surrounding tissue that determine whether they divide, differentiate or die. Consistent with this model, deregulation of signalling pathways is frequently associated with human cancer, which is characterised by the uncontrolled growth of cells. Indeed, new therapeutic approaches in cancer therapy, such as Herceptin, target signalling pathway components.
Most oncogenic pathways have been highly conserved throughout evolution and of the known cancer genes, over 90 % are present in Drosophila. These cancer pathways include Wnt, Notch, Ras, Hippo, PI3K and JAK/STAT signalling cascades. Drosophila represents a particularly powerful genetic model for the analysis of signalling cascades. Forward genetic screens in Drosophila have identified numerous pathway components later described as tumour suppressors and oncogenes in human. The dissection of signalling pathways in Drosophila is facilitated by the availability of a broad range of genetic tools, a completely sequenced genome and the availability of genome-wide collections of RNA interference (RNAi) reagents and high-throughput methods in cells and in vivo.
Project objectives
In the European Union (EU) Seventh Framework Programme (FP7) project CANCERPATHWAYS we set out to identify novel targets and small molecules using Drosophila as model organism. Our strategy was based on high-throughput genome-wide RNAi screens in cultured Drosophila cells in vitro and Drosophila flies in vivo. The recent discovery of RNAi allows the systematic knockdown of all genes in an organism. RNAi experiments in Drosophila are facilitated by a low genetic redundancy and highly efficient target gene knockdown by long double-stranded RNAs (dsRNAs). Testing of small molecular compounds for their potency to modify signalling pathway activity is supposed to afford the identification of potential novel therapeutic agents.
In order to perform screenings for targets and molecules in a high-throughput format, robust in vitro and in vivo bioassays are necessary. The resulting data have to be standardised and integrated with powerful bioinformatics tools. Potential targets and drug-like substances have to be validated in the developed in vivo assays for their mode of action. The information obtained during the project will be made publicly available and disseminated in databases.
The CANCERPATHWAYS project integrated recent technological advances, such as genome-wide RNAi libraries for cell-based and in vivo assay models that were developed by the participants, as well as computational approaches and databases to integrate and disseminate information obtained during the project.
The specific objectives of the CANCERPATHWAYS project were:
1. establishment of high-throughput bioassays for major oncogenic pathways for cell-based and in vivo screening approaches
2. identification of novel targets of oncogenic signalling pathways by RNAi screening in cultured cells and in vivo
3. identification of compounds that modify oncogenic signalling pathways by screening in cultured cells and in vivo
Project results:
The CANCERPATHWAYS project consisted of five connected activities:
(1) establishment of high-throughput bioassays in cultured Drosophila cells in vitro and Drosophila flies in vivo,
(2) cell-based screening,
(3) screening in Drosophila flies,
(4) in vivo validation of identified targets and compounds and
(5) analysis, integration and dissemination of obtained results.
These activities included the design and generation of novel RNAi libraries for cell-based and in vivo RNAi screening in Drosophila. For the analysis and integration of the obtained results, bioinformatics tools were developed.
Genome-wide RNAi libraries
RNAi has proven to be a valid tool in Drosophila to systematically identify signalling pathway components. However, RNAi experiments have inherent false-positive and false-negative rates. Thus, we set out to use in silico tools to predict, and hence avoid, the majorities of these effects. We have developed a novel software algorithm, which can be used for the design and evaluation of genome-wide RNAi libraries. This software, NEXT-RNAi, is publically available at http://www.nextrnai.org(si apre in una nuova finestra) A second-generation Drosophila dsRNA library was designed using the software NEXT-RNAi and synthesised by a two-step PCR approach followed by in vitro transcription. The library is available for the CANCERPATHWAYS consortium and external collaboration partners in the Boutros lab (see http://www.dkfz.de/signaling(si apre in una nuova finestra) for details) and in the Sheffield RNAi Screening Facility (see http://www.shef.ac.uk/bms/research/rnai(si apre in una nuova finestra) for details). It has been distributed to researchers in Europe and worldwide.
Similar problems with false-positive and false-negative rates held true for the first in vivo RNAi Drosophila library as well. Within the CANCERPATHWAYS project we have also generated a second-generation in vivo Drosophila RNAi library using the novel Drosophila dsRNA library described above as basics. Furthermore, this library has been created using the novel method of site-specific transgenesis, mediated by phiC31 integrase. A suitable 'landing site' for integration has been identified by various test experiments. Transgenic Drosophila lines have been created covering about 12 000 genes up to now. All lines from this library (named 'KK') are publically available through the Vienna Drosophila RNAi Center, VDRC (see http://www.vdrc.at(si apre in una nuova finestra) for details). The VDRC has already shipped more than 200 000 lines from this library to Drosophila researchers in Europe and worldwide.
Bioinformatic tools for data analysis
High-throughput assays generate large datasets that need to be processed in order to identify specific hits for both, RNAi and compound screening experiments. Previously we had released the improved software tool cellHTS2 for the analysis of genome-wide RNAi screens. We further improved analysis methods and regularly released novel cellHTS2 software versions to the public as part of Bioconductor (see http://www.bioconductor.org(si apre in una nuova finestra) for details). So far, the software has found many users in academia and industry, reflected by more than 2000 software downloads per year. This software is also available as web-application for the user-friendly analysis of single- and multi-channel experiments, which facilitates the analysis of high-throughput RNAi and compound data sets (see http://www.web-cellHTS2.dkfz.de(si apre in una nuova finestra) for details), developed by the CANCERPATHWAYS project. web-cell THS2 is widely used by researchers worldwide and has been visited about 15 000 times during the past years.
In the CANCERPATHWAYS project, we also performed large-scale RNAi screens with imaging readout. Image-based screening allows the analysis of multiparametric features instead of one feature only like pathway activity or cell proliferation. However, it also has put new and substantial requirements on data management and analysis. These have been addressed by the imageHTS and EBImage packages for R we have developed and which are distributed on Bioconductor (see http://www.bioconductor.org(si apre in una nuova finestra) for details). These have been heavily used by the research community as reflected by about 1000 and 6000 software downloads per year, respectively.
We and other researchers worldwide performed cell-based and in vivo RNAi screens in Drosophila and other species. In order to publically provide and efficiently compare screening results we have further developed the RNAi databases GenomeRNAi (see http://www.genomernai.org/(si apre in una nuova finestra) for details) and FLIGHT (see http://flight.icr.ac.uk/(si apre in una nuova finestra) for details) with new functionalities. Both databases are also accessible via a common portal, which we have developed as part of CANCERPATHWAYS (see http://www.cancerpathways.eu/SearchGeneRNAi/(si apre in una nuova finestra) for details). In addition to the published phenotypes from RNAi screens, the databases provide updated resources of RNAi reagents and their predicted qualities that are available for the Drosophila and the human genome. The databases connect observed phenotypes with annotations of the targeted gene and information about the RNAi reagent used for the perturbation experiment. Large-scale RNAi screens performed as part of the CANCERPATHWAYS project were and will be made publically available after publication.
RNAi and compound screening
We have established a comprehensive set of bioassays for diverse oncogenic signalling pathways. Assay protocols for cell-based and in vivo approaches were adapted to high-throughput and high-content screening conditions. Novel assays were developed for screening of small molecular compound libraries for their impact on cell fitness and specific pathway activities. In particular, in vivo models were established to monitor pathway activity, tumour development and metastasis in Drosophila. After successful validation with a defined set of positive and negative controls, these assays were applied to perform high-throughput cell-based and in vivo RNAi and compound screens in Drosophila.
High-throughput RNAi screening resulted in the identification of several novel candidates involved in cancer-relevant signalling pathways and processes. These candidates were validated and their roles in the respective signalling processes determined. In the following, published examples of our work are outlined, demonstrating the impact on high-throughput RNAi screening in vivo and in cells. Further results from the CANCERPATHWAYS research are already under revision in international journals or are currently pursued to further characterise the role of selected candidates.
The JAK/STAT pathway has been identified as an important mediator of tumourigenesis in a large number of human cancers and participates in the regulation of malignant processes. In addition to roles in multiple solid tumours, a key role is played in the development of a wide range of haematopoetic tumours. By RNAi screening, we identified transcriptional targets of Drosophila JAK/STAT pathway signalling as effectors of haematopoietic tumour formation.
Notch signalling plays a key role in the development of organisms and its deregulation has been linked to cancer. We have performed large-scale RNAi screening and identified novel mechanisms that control Notch signalling. We showed that JAK/STAT and four-jointed, both known to be key mediators in other signalling pathways, play a role in the position and function of the Notch-mediated eye growth organiser. In another study, Notch signalling was further shown to be regulated by the conserved miR-8/200 microRNA family in development and cancer cells. RNAi screening on Notch signalling lead to the finding that the histone demethylase KDM5A is an integral part of the core Notch-RBP-J repressor complex.
Mutations in the RAS family of proto-oncogenes are very common, being found in 20 to 30 % of all human tumours. Our studies on Drosophila RAS/MAPK signalling indicate a role in the regulation of innate immune responses in immune and intestinal stem cells. Our observations provide an example of a pathway that promotes cell proliferation and has simultaneously been utilised to limit the immune response. We also carried out a project to map signalling networks through synthetic genetic interaction analysis using RNAi. Computational analysis of this interaction matrix allowed us to reconstruct signalling pathways and identify a novel conserved regulator of Ras-MAPK signalling.
The Hippo pathway was uncovered as a key regulator of tissue growth in Drosophila. Like the Insulin/PI3K pathway, the Hippo pathway can drive tissue growth without influencing cell differentiation. In an in vivo RNAi approach we identified Kibra as a regulator of the Hippo signalling network and further characterised its role in this pathway.
Cell division is accompanied by dynamic changes in cell shape. Disruption of this process may lead to uncontrolled cell division and, thus, cancer. Our studies have built on our previous work in Drosophila cell culture showing a critical role for the protein Moesin in mitotic cell rounding. Significantly, our studies revealed a role for Moesin-induced changes in shape and in cells dividing in unstructured environments, which we used as a model for metastasis. This is likely to be relevant to cancer as one of the human homologues of Moesin, Ezrin, is upregulated in metastatic cancers. We published data from a genome-wide RNAi screen in Drosophila cell culture for genes and pathways that regulate cell shape. This identified a large number of novel regulators, which were then analysed in a parallel screen in two human cell lines.
The development of preclinical models suitable for live animal small compound screenings is needed urgently to speed the pre-screen of new drugs for possible use in the clinic. Screenings in mice are time-consuming and expensive and permit only low throughput. Based on their ease of handling and cultivation, we used Drosophila melanogaster cancer paradigms to assess their suitability for in vivo drug discovery.
We performed large-scale screens in Drosophila cells and in vivo to identify small molecules that modify oncogenic signalling pathways. Hits were selected and validated in independent experiments. Moreover, some of the drugs exhibited a dose-dependent effect. In addition to known anti-cancer drugs that were part of the chemical libraries, novel compounds were identified and their mode of action followed up. Results are expected to lead to patentable knowledge.
Our data on in vivo compound screening in Drosophila provide the proof-of-principle of a whole organism-based in vivo drug screening strategy that is at the same time low cost and allows high-throughput screens. The majority of the drugs identified during the course of the screen appear to represent high therapeutic value, and all together these data strongly validate that our Drosophila cancer paradigm applied is extremely powerful for primary in vivo drug screening experiments.
Dissemination of results
Dissemination of knowledge generated within the CANCERPATHWAYS project has been realised by presenting results at international scientific meetings and workshops. Several manuscripts have been published or are in preparation for publication. RNAi screening and bioinformatics training courses were organised by CANCERPATHWAYS members for transfer of knowledge. Reagents generated during the course of the project were made available to the Drosophila community as described above.
The CANCERPATHWAYS project website (see http://www.cancerpathways.eu(si apre in una nuova finestra) online) displays information about the project, it aims and the consortium members. It contains several sites including relevant publications, events, links, trainings, job offers and articles written to enhance the general understanding of our research.
We have further designed and distributed the brochure 'Drosophila as a model system to study cancer' to increase the public understanding about the importance of model organisms for cancer research. A downloadable version is available on our website at http://www.cancerpathways.eu/Brochure(si apre in una nuova finestra)
Potential impact:
CANCERPATHWAYS brought together a strong team of researchers with complementary expertise in Drosophila RNAi, pathway analysis, target identification and bioinformatics from across Europe. The approach taken in this collaborative project integrated and compared state-of-the-art assays on most known signalling pathways involved in human cancer in the Drosophila model system. It further integrated the data generated by innovative and standardised computational analysis.
The project afforded the establishment of in vitro and in vivo models of signalling pathways for high-throughput screening and to test the activity of drugs using developmental and cancer phenotypes. During this process, we integrated recent technological advances, such as genome-wide RNAi libraries for cell-based assays and in vivo models and generated novel tools that have a significant impact on the field far beyond the scope of this project. The RNAi resources generated as part of the CANCERPATHWAYS project are available for the research community and have been distributed to researchers in Europe and worldwide. The genome-wide library for cell-based RNAi screening in Drosophila is available in the Boutros lab (see http://www.dkfz.de/signaling(si apre in una nuova finestra) for details) and in the Sheffield RNAi Screening Facility (see http://www.shef.ac.uk/bms/research/rnai(si apre in una nuova finestra) for details). All Drosophila lines from the in vivo RNAi library (named 'KK') are publically available through the Vienna Drosophila RNAi Center, VDRC (see http://www.vdrc.at(si apre in una nuova finestra) for details).
One important point of the project was the strong integration of bioinformatics tools with the screening procedure. These tools have been further developed within this project. They are of immense use to the scientific community and were made freely available. The bioinformatic tools include the data analysis software cellHTS2, imageHTS and EBImage, all available as Bioconductor packages in R (see http://www.bioconductor.org(si apre in una nuova finestra) for details), and web-cellHTS2 (see http://www.web-cellHTS2.dkfz.de(si apre in una nuova finestra) for details). The tools further include the RNAi design software NEXT-RNAi (see http://www.nextrnai.org(si apre in una nuova finestra) for details) and the RNAi databases GenomeRNAi (see http://www.genomernai.org(si apre in una nuova finestra) for details) and FLIGHT (see http://www.flight.icr.ac.uk/(si apre in una nuova finestra) for details) for those a common access portal has been developed (see http://www.cancerpathways.eu/SearchGeneRNAi/(si apre in una nuova finestra) for details).
Using cell-based and in vivo RNAi screening approaches we have identified multiple novel key factors involved in aberrant cancer that would not otherwise been identified by traditional 'top down' approaches. Knowledge on signalling pathways generated within this project strongly contributed to a better understanding of fundamental cellular processes and, thus, is of value in cancer research and many other research areas. The parallel screening for drug-like substances in the assay systems additionally provided small molecules candidates. Further research will be carried out by the CANCERPATHWAYS partners to characterise small molecular compounds in depth and evaluate their potential for the development of novel anti-cancer drugs. It is expected that the results obtained will generate patents even after the end of the CANCERPATHWAYS project.
The depth and breadth of the consortium enabled the consortium to have a substantial impact on developing the essential knowledge and platform technologies in high-throughput pathway analysis for rapid identification of cancer targets. This is of the benefit for the European Research Area and, indeed, the scientific community worldwide. The combination of expertise of the participants is unique and the target discovery efforts through RNAi in vivo and in cell-based high-throughput assays unparalleled worldwide. The research programme of this consortium has strengthened functional genomics research in Europe through providing tools and strategies that are widely applicable for target identification far beyond the field of cancer research.
Project website:
The project website contains several sites with information about relevant publications, events, links, trainings and job offers and includes a restricted area dedicated to information exchange between the participants. We have created the brochure 'Drosophila as a model system to study cancer' to increase the public understanding about the importance of model organisms for cancer research. A downloadable version of the brochure can be found on our website at http://www.cancerpathways.eu/Brochure/(si apre in una nuova finestra)
CANCERPATHWAYS contact:
Coordinator
Prof. Dr. Michael Boutros
German Cancer Research Center (DKFZ) & University of Heidelberg
Scientific project management
Dr Ulrike Hardeland
Email: u.hardeland@dkfz.de
Phone: +49-622-1421961
Fax: +49-622-1421959
Postal address
German Cancer Research Center (DKFZ)
Im Neuenheimer Feld 580
69120 Heidelberg
Germany
Tumour development is characterised by the uncontrolled growth of cells. The deregulation of diverse signal transduction pathways has previously been linked with carcinogenesis. These cancer pathways include Wnt, Notch, Ras, Hippo, PI3K and JAK/STAT signalling cascades. Although new therapeutic approaches in cancer treatment target signalling pathway components, the full potential of signalling molecules as therapy targets remains to be explored. The CANCERPATHWAYS consortium aimed to identify novel targets and drug-like molecules for therapeutic application. In this respect we made use of the high evolutionary conservation of most oncogenic pathways with Drosophila representing a particularly powerful model system for the analysis of signalling cascades.
The CANCERPATHWAYS project integrated recent technological advances, such as genome-wide RNAi and compound screens in cell-based and in vivo models as well as computational approaches. We have established a comprehensive set of cell-based and in vivo bioassays for diverse oncogenic signalling pathways. Novel assays were developed for screening of small molecular compound libraries for their impact on cell fitness, specific pathway activities, tumour growth and metastasis in Drosophila. These assays were successfully validated and applied to perform high-throughput cell-based and in vivo RNAi and compound screens in Drosophila.
RNAi has proven to be a valid tool in Drosophila to systematically identify signalling pathway components. However, experiments with previous RNAi libraries had inherent false-positive and false-negative rates. Thus, we computationally designed and synthesised novel genome-wide RNAi libraries for cell-based and in vivo RNAi screening in Drosophila. Furthermore, high-throughput assays generate large datasets that need to be processed in order to identify specific hits for both, RNAi and compound screening experiments. One important point of the project was the strong integration of bioinformatics tools with the screening procedure. We have developed and improved data analysis tools and databases displaying the screening results. All these resources are of immense use and were made available for the research community.
Applying the technological advancements of the CANCERPATHWAYS project, we have identified several novel candidate genes and demonstrated their function in cancer signalling pathways. Many of those studies are already published by the CANCERPATHWAYS consortium while further studies will be published soon. High-throughput compound screens in Drosophila cells and in vivo tumour models resulted in the identification of known anti-cancer drugs as well as novel compounds that interfere with the activity of oncogenic signalling pathways in vitro and in vivo models.
Project context and objectives:
Cancer is one of the leading causes of death in Europe and worldwide. Although significant advances have been made in cancer research, it still remains a public health concern. It is therefore of great importance to develop new and specific drug treatments for cancer. The EU medium scale focused research project CANCERPATHWAYS combines scientific expertise of eight European partners to produce fundamental advances in our understanding of cancer biology and how this disease can be attacked.
Drosophila as a model system to study cancer
Cancer has often been characterised as a disease of signal transduction pathways. During development and homeostasis, cells constantly receive signals from surrounding tissue that determine whether they divide, differentiate or die. Consistent with this model, deregulation of signalling pathways is frequently associated with human cancer, which is characterised by the uncontrolled growth of cells. Indeed, new therapeutic approaches in cancer therapy, such as Herceptin, target signalling pathway components.
Most oncogenic pathways have been highly conserved throughout evolution and of the known cancer genes, over 90 % are present in Drosophila. These cancer pathways include Wnt, Notch, Ras, Hippo, PI3K and JAK/STAT signalling cascades. Drosophila represents a particularly powerful genetic model for the analysis of signalling cascades. Forward genetic screens in Drosophila have identified numerous pathway components later described as tumour suppressors and oncogenes in human. The dissection of signalling pathways in Drosophila is facilitated by the availability of a broad range of genetic tools, a completely sequenced genome and the availability of genome-wide collections of RNA interference (RNAi) reagents and high-throughput methods in cells and in vivo.
Project objectives
In the European Union (EU) Seventh Framework Programme (FP7) project CANCERPATHWAYS we set out to identify novel targets and small molecules using Drosophila as model organism. Our strategy was based on high-throughput genome-wide RNAi screens in cultured Drosophila cells in vitro and Drosophila flies in vivo. The recent discovery of RNAi allows the systematic knockdown of all genes in an organism. RNAi experiments in Drosophila are facilitated by a low genetic redundancy and highly efficient target gene knockdown by long double-stranded RNAs (dsRNAs). Testing of small molecular compounds for their potency to modify signalling pathway activity is supposed to afford the identification of potential novel therapeutic agents.
In order to perform screenings for targets and molecules in a high-throughput format, robust in vitro and in vivo bioassays are necessary. The resulting data have to be standardised and integrated with powerful bioinformatics tools. Potential targets and drug-like substances have to be validated in the developed in vivo assays for their mode of action. The information obtained during the project will be made publicly available and disseminated in databases.
The CANCERPATHWAYS project integrated recent technological advances, such as genome-wide RNAi libraries for cell-based and in vivo assay models that were developed by the participants, as well as computational approaches and databases to integrate and disseminate information obtained during the project.
The specific objectives of the CANCERPATHWAYS project were:
1. establishment of high-throughput bioassays for major oncogenic pathways for cell-based and in vivo screening approaches
2. identification of novel targets of oncogenic signalling pathways by RNAi screening in cultured cells and in vivo
3. identification of compounds that modify oncogenic signalling pathways by screening in cultured cells and in vivo
Project results:
The CANCERPATHWAYS project consisted of five connected activities:
(1) establishment of high-throughput bioassays in cultured Drosophila cells in vitro and Drosophila flies in vivo,
(2) cell-based screening,
(3) screening in Drosophila flies,
(4) in vivo validation of identified targets and compounds and
(5) analysis, integration and dissemination of obtained results.
These activities included the design and generation of novel RNAi libraries for cell-based and in vivo RNAi screening in Drosophila. For the analysis and integration of the obtained results, bioinformatics tools were developed.
Genome-wide RNAi libraries
RNAi has proven to be a valid tool in Drosophila to systematically identify signalling pathway components. However, RNAi experiments have inherent false-positive and false-negative rates. Thus, we set out to use in silico tools to predict, and hence avoid, the majorities of these effects. We have developed a novel software algorithm, which can be used for the design and evaluation of genome-wide RNAi libraries. This software, NEXT-RNAi, is publically available at http://www.nextrnai.org(si apre in una nuova finestra) A second-generation Drosophila dsRNA library was designed using the software NEXT-RNAi and synthesised by a two-step PCR approach followed by in vitro transcription. The library is available for the CANCERPATHWAYS consortium and external collaboration partners in the Boutros lab (see http://www.dkfz.de/signaling(si apre in una nuova finestra) for details) and in the Sheffield RNAi Screening Facility (see http://www.shef.ac.uk/bms/research/rnai(si apre in una nuova finestra) for details). It has been distributed to researchers in Europe and worldwide.
Similar problems with false-positive and false-negative rates held true for the first in vivo RNAi Drosophila library as well. Within the CANCERPATHWAYS project we have also generated a second-generation in vivo Drosophila RNAi library using the novel Drosophila dsRNA library described above as basics. Furthermore, this library has been created using the novel method of site-specific transgenesis, mediated by phiC31 integrase. A suitable 'landing site' for integration has been identified by various test experiments. Transgenic Drosophila lines have been created covering about 12 000 genes up to now. All lines from this library (named 'KK') are publically available through the Vienna Drosophila RNAi Center, VDRC (see http://www.vdrc.at(si apre in una nuova finestra) for details). The VDRC has already shipped more than 200 000 lines from this library to Drosophila researchers in Europe and worldwide.
Bioinformatic tools for data analysis
High-throughput assays generate large datasets that need to be processed in order to identify specific hits for both, RNAi and compound screening experiments. Previously we had released the improved software tool cellHTS2 for the analysis of genome-wide RNAi screens. We further improved analysis methods and regularly released novel cellHTS2 software versions to the public as part of Bioconductor (see http://www.bioconductor.org(si apre in una nuova finestra) for details). So far, the software has found many users in academia and industry, reflected by more than 2000 software downloads per year. This software is also available as web-application for the user-friendly analysis of single- and multi-channel experiments, which facilitates the analysis of high-throughput RNAi and compound data sets (see http://www.web-cellHTS2.dkfz.de(si apre in una nuova finestra) for details), developed by the CANCERPATHWAYS project. web-cell THS2 is widely used by researchers worldwide and has been visited about 15 000 times during the past years.
In the CANCERPATHWAYS project, we also performed large-scale RNAi screens with imaging readout. Image-based screening allows the analysis of multiparametric features instead of one feature only like pathway activity or cell proliferation. However, it also has put new and substantial requirements on data management and analysis. These have been addressed by the imageHTS and EBImage packages for R we have developed and which are distributed on Bioconductor (see http://www.bioconductor.org(si apre in una nuova finestra) for details). These have been heavily used by the research community as reflected by about 1000 and 6000 software downloads per year, respectively.
We and other researchers worldwide performed cell-based and in vivo RNAi screens in Drosophila and other species. In order to publically provide and efficiently compare screening results we have further developed the RNAi databases GenomeRNAi (see http://www.genomernai.org/(si apre in una nuova finestra) for details) and FLIGHT (see http://flight.icr.ac.uk/(si apre in una nuova finestra) for details) with new functionalities. Both databases are also accessible via a common portal, which we have developed as part of CANCERPATHWAYS (see http://www.cancerpathways.eu/SearchGeneRNAi/(si apre in una nuova finestra) for details). In addition to the published phenotypes from RNAi screens, the databases provide updated resources of RNAi reagents and their predicted qualities that are available for the Drosophila and the human genome. The databases connect observed phenotypes with annotations of the targeted gene and information about the RNAi reagent used for the perturbation experiment. Large-scale RNAi screens performed as part of the CANCERPATHWAYS project were and will be made publically available after publication.
RNAi and compound screening
We have established a comprehensive set of bioassays for diverse oncogenic signalling pathways. Assay protocols for cell-based and in vivo approaches were adapted to high-throughput and high-content screening conditions. Novel assays were developed for screening of small molecular compound libraries for their impact on cell fitness and specific pathway activities. In particular, in vivo models were established to monitor pathway activity, tumour development and metastasis in Drosophila. After successful validation with a defined set of positive and negative controls, these assays were applied to perform high-throughput cell-based and in vivo RNAi and compound screens in Drosophila.
High-throughput RNAi screening resulted in the identification of several novel candidates involved in cancer-relevant signalling pathways and processes. These candidates were validated and their roles in the respective signalling processes determined. In the following, published examples of our work are outlined, demonstrating the impact on high-throughput RNAi screening in vivo and in cells. Further results from the CANCERPATHWAYS research are already under revision in international journals or are currently pursued to further characterise the role of selected candidates.
The JAK/STAT pathway has been identified as an important mediator of tumourigenesis in a large number of human cancers and participates in the regulation of malignant processes. In addition to roles in multiple solid tumours, a key role is played in the development of a wide range of haematopoetic tumours. By RNAi screening, we identified transcriptional targets of Drosophila JAK/STAT pathway signalling as effectors of haematopoietic tumour formation.
Notch signalling plays a key role in the development of organisms and its deregulation has been linked to cancer. We have performed large-scale RNAi screening and identified novel mechanisms that control Notch signalling. We showed that JAK/STAT and four-jointed, both known to be key mediators in other signalling pathways, play a role in the position and function of the Notch-mediated eye growth organiser. In another study, Notch signalling was further shown to be regulated by the conserved miR-8/200 microRNA family in development and cancer cells. RNAi screening on Notch signalling lead to the finding that the histone demethylase KDM5A is an integral part of the core Notch-RBP-J repressor complex.
Mutations in the RAS family of proto-oncogenes are very common, being found in 20 to 30 % of all human tumours. Our studies on Drosophila RAS/MAPK signalling indicate a role in the regulation of innate immune responses in immune and intestinal stem cells. Our observations provide an example of a pathway that promotes cell proliferation and has simultaneously been utilised to limit the immune response. We also carried out a project to map signalling networks through synthetic genetic interaction analysis using RNAi. Computational analysis of this interaction matrix allowed us to reconstruct signalling pathways and identify a novel conserved regulator of Ras-MAPK signalling.
The Hippo pathway was uncovered as a key regulator of tissue growth in Drosophila. Like the Insulin/PI3K pathway, the Hippo pathway can drive tissue growth without influencing cell differentiation. In an in vivo RNAi approach we identified Kibra as a regulator of the Hippo signalling network and further characterised its role in this pathway.
Cell division is accompanied by dynamic changes in cell shape. Disruption of this process may lead to uncontrolled cell division and, thus, cancer. Our studies have built on our previous work in Drosophila cell culture showing a critical role for the protein Moesin in mitotic cell rounding. Significantly, our studies revealed a role for Moesin-induced changes in shape and in cells dividing in unstructured environments, which we used as a model for metastasis. This is likely to be relevant to cancer as one of the human homologues of Moesin, Ezrin, is upregulated in metastatic cancers. We published data from a genome-wide RNAi screen in Drosophila cell culture for genes and pathways that regulate cell shape. This identified a large number of novel regulators, which were then analysed in a parallel screen in two human cell lines.
The development of preclinical models suitable for live animal small compound screenings is needed urgently to speed the pre-screen of new drugs for possible use in the clinic. Screenings in mice are time-consuming and expensive and permit only low throughput. Based on their ease of handling and cultivation, we used Drosophila melanogaster cancer paradigms to assess their suitability for in vivo drug discovery.
We performed large-scale screens in Drosophila cells and in vivo to identify small molecules that modify oncogenic signalling pathways. Hits were selected and validated in independent experiments. Moreover, some of the drugs exhibited a dose-dependent effect. In addition to known anti-cancer drugs that were part of the chemical libraries, novel compounds were identified and their mode of action followed up. Results are expected to lead to patentable knowledge.
Our data on in vivo compound screening in Drosophila provide the proof-of-principle of a whole organism-based in vivo drug screening strategy that is at the same time low cost and allows high-throughput screens. The majority of the drugs identified during the course of the screen appear to represent high therapeutic value, and all together these data strongly validate that our Drosophila cancer paradigm applied is extremely powerful for primary in vivo drug screening experiments.
Dissemination of results
Dissemination of knowledge generated within the CANCERPATHWAYS project has been realised by presenting results at international scientific meetings and workshops. Several manuscripts have been published or are in preparation for publication. RNAi screening and bioinformatics training courses were organised by CANCERPATHWAYS members for transfer of knowledge. Reagents generated during the course of the project were made available to the Drosophila community as described above.
The CANCERPATHWAYS project website (see http://www.cancerpathways.eu(si apre in una nuova finestra) online) displays information about the project, it aims and the consortium members. It contains several sites including relevant publications, events, links, trainings, job offers and articles written to enhance the general understanding of our research.
We have further designed and distributed the brochure 'Drosophila as a model system to study cancer' to increase the public understanding about the importance of model organisms for cancer research. A downloadable version is available on our website at http://www.cancerpathways.eu/Brochure(si apre in una nuova finestra)
Potential impact:
CANCERPATHWAYS brought together a strong team of researchers with complementary expertise in Drosophila RNAi, pathway analysis, target identification and bioinformatics from across Europe. The approach taken in this collaborative project integrated and compared state-of-the-art assays on most known signalling pathways involved in human cancer in the Drosophila model system. It further integrated the data generated by innovative and standardised computational analysis.
The project afforded the establishment of in vitro and in vivo models of signalling pathways for high-throughput screening and to test the activity of drugs using developmental and cancer phenotypes. During this process, we integrated recent technological advances, such as genome-wide RNAi libraries for cell-based assays and in vivo models and generated novel tools that have a significant impact on the field far beyond the scope of this project. The RNAi resources generated as part of the CANCERPATHWAYS project are available for the research community and have been distributed to researchers in Europe and worldwide. The genome-wide library for cell-based RNAi screening in Drosophila is available in the Boutros lab (see http://www.dkfz.de/signaling(si apre in una nuova finestra) for details) and in the Sheffield RNAi Screening Facility (see http://www.shef.ac.uk/bms/research/rnai(si apre in una nuova finestra) for details). All Drosophila lines from the in vivo RNAi library (named 'KK') are publically available through the Vienna Drosophila RNAi Center, VDRC (see http://www.vdrc.at(si apre in una nuova finestra) for details).
One important point of the project was the strong integration of bioinformatics tools with the screening procedure. These tools have been further developed within this project. They are of immense use to the scientific community and were made freely available. The bioinformatic tools include the data analysis software cellHTS2, imageHTS and EBImage, all available as Bioconductor packages in R (see http://www.bioconductor.org(si apre in una nuova finestra) for details), and web-cellHTS2 (see http://www.web-cellHTS2.dkfz.de(si apre in una nuova finestra) for details). The tools further include the RNAi design software NEXT-RNAi (see http://www.nextrnai.org(si apre in una nuova finestra) for details) and the RNAi databases GenomeRNAi (see http://www.genomernai.org(si apre in una nuova finestra) for details) and FLIGHT (see http://www.flight.icr.ac.uk/(si apre in una nuova finestra) for details) for those a common access portal has been developed (see http://www.cancerpathways.eu/SearchGeneRNAi/(si apre in una nuova finestra) for details).
Using cell-based and in vivo RNAi screening approaches we have identified multiple novel key factors involved in aberrant cancer that would not otherwise been identified by traditional 'top down' approaches. Knowledge on signalling pathways generated within this project strongly contributed to a better understanding of fundamental cellular processes and, thus, is of value in cancer research and many other research areas. The parallel screening for drug-like substances in the assay systems additionally provided small molecules candidates. Further research will be carried out by the CANCERPATHWAYS partners to characterise small molecular compounds in depth and evaluate their potential for the development of novel anti-cancer drugs. It is expected that the results obtained will generate patents even after the end of the CANCERPATHWAYS project.
The depth and breadth of the consortium enabled the consortium to have a substantial impact on developing the essential knowledge and platform technologies in high-throughput pathway analysis for rapid identification of cancer targets. This is of the benefit for the European Research Area and, indeed, the scientific community worldwide. The combination of expertise of the participants is unique and the target discovery efforts through RNAi in vivo and in cell-based high-throughput assays unparalleled worldwide. The research programme of this consortium has strengthened functional genomics research in Europe through providing tools and strategies that are widely applicable for target identification far beyond the field of cancer research.
Project website:
The project website contains several sites with information about relevant publications, events, links, trainings and job offers and includes a restricted area dedicated to information exchange between the participants. We have created the brochure 'Drosophila as a model system to study cancer' to increase the public understanding about the importance of model organisms for cancer research. A downloadable version of the brochure can be found on our website at http://www.cancerpathways.eu/Brochure/(si apre in una nuova finestra)
CANCERPATHWAYS contact:
Coordinator
Prof. Dr. Michael Boutros
German Cancer Research Center (DKFZ) & University of Heidelberg
Scientific project management
Dr Ulrike Hardeland
Email: u.hardeland@dkfz.de
Phone: +49-622-1421961
Fax: +49-622-1421959
Postal address
German Cancer Research Center (DKFZ)
Im Neuenheimer Feld 580
69120 Heidelberg
Germany
