for SLAS High-Content Screening Conference 2017
View analytic
Wednesday, September 20 • 12:00pm - 2:30pm
Session 4: Model Systems
The selection of a suitable model system is critical in the successful advancement of novel therapeutic approaches to disease. The development of models and assays that recapitulate the full complexity of the disease they intend to represent, and the ability to scale them up to screening campaigns remains a big challenge in the field. The session will bring together experts in the field presenting innovative models across different disease areas. 

12:00 - 12:30 
Presentation Title Forthcoming
Marc Mercola, Stanford University 

12:30 - 1:00 
Combining functional genomics screening, 3D models and multiparametric phenotypic analysis supports identification of selected ARF GTPases as modulators of polarity and invasion.
Emma Shanks, Cancer Research UK Beatson Institute

The utility of 3D models in high throughput screening, high content imaging and phenotypic analysis is currently a topic of much discussion. Here we present data from a screen investigating ARF GTPases, regulators and effectors in a prostate cancer spheroid model. We conducted a screen using a bespoke library of shRNAs which co-expressed a fluorescent tag to allow for intrawell phenotype normalisation between infected and non-infected spheres. Multiparametric phenotypic image analysis was utilised to identify, teach and quantify phenotypes pertaining to polarity, in conjunction with 'standard' morphological analysis. All screening was conducted in 2D and 3D, with the latter evaluated using fixed and live cell imaging. Informatics was employed to introduce streamlined data handling workflows, based principally on KNIME, to allow interpretation of data-rich phenotypic analysis. Target deconvolution and validation approaches identified specific ARF GTPases as modulators of polarity and potential therapeutic targets for metastatic disease. We have conducted i) 2D vs 3D and ii) 3D fixed vs live cell inter-platform comparisons. This has provided insights into the value of each model system, but moreover, has allowed us to explore the phenotypic consistencies between models, and therefore their predictive power. In particular, we identified a combination of phenotypes which, when present in 3D fixed cells, successfully predict invasive potential, which we validated using longer-term, live cell assays. These findings will be discussed.

1:00 - 1:30 
Presentation Title Forthcoming 
Christian Helker, Max Planck Institute

Historically, screens to identify small molecules are done through target and phenotypic based in vitro screens.  However, the identified drugs often fail in the preclinical animal model for two main reasons: 1) pharmacokinetics at the whole organism level, including the absorption, distribution, metabolism, and excretion (“ADME”) properties of small molecules2) cytotoxicity and side/off target effects.  These complexities hinder the extrapolation of drug-target interactions in vitro to pharmacological actions in vivo, resulting in more than 70% of compounds in oncology to fail in phase II clinical trials, while 59% of the remaining compounds are discarded in phase III due to intolerable toxicities.Considering these challenges, developing tools for rapid, cost-efficient and translational small-molecule discovery in whole organisms greatly strengthens the translation potential in novel drug discovery. Zebrafish embryos are amenable to semi high-throughput chemical screens combining the advantages of the scale of in vitro screens with the physiological complexity of a developing animal. Using the zebrafish as an in vivo system for drug discovery has several unique advantages: (i) the low cost per animal, (ii) ease of genetic manipulations  (iii) small body size enables to screen zebrafish larvae in 96-well plates (iv) large progeny of embryos to ensure high throughput of large chemical libraries (v) as a vertebrate, the zebrafish shares a similar body plan to mammals and genetic and molecular pathways that drive organ development are conserved to humans Altogether, this brings the zebrafish in a prominent position as an affordable and feasible model system to perform high throughput in vivo screens.  

1:30 - 2:00 
Collaborative Phenotyping at King's College London: HipSci and the Stem Cell Hotel
Davide Danovi, King's College London, Director, HipSci Cell Phenotyping

A clear understanding of the limitation of the current approaches in Drug Discovery is emerging throughout academia and industry. Efforts incorporating relevant cells, advanced cell culture techniques, artificial microenvironment, imaging and data analysis are moving forward allowing unprecedented opportunities in the modelling of diseases and identification of targets and therapies. The HipSci project brings together the Wellcome Trust Sanger Institute, the University of Dundee, King’s College London and the European Bioinformatics institute. A large panel of hundreds of cell lines reprogrammed from adult cells donated by healthy volunteers and patients are being fully characterised in genomics, proteomics and cell behaviour. We work within this framework at the Centre for Stem Cells and Regenerative Medicine directed by Fiona Watt at King’s College London. We have established assays for imaging of induced pluripotent stem cells and are also developing solutions to integrate dynamic and end-point high content data with the other datasets provided by the partner centres in the project. The expertise in stem cell biology, image analysis and engineered substrates as well as access to specialised equipment are now opening to interested scientists through a dedicated collaborative phenotyping space, the Stem Cell Hotel.

2:00 - 2:30 
Phenomics in the precision medicine of solid tumors
Vilja Pietiäinen, Institute for Molecular Medicine Finland FIMM

Our aim is to gain better understanding of the cancer biology and eventually to tailor drugs for patients by extensive -omics profiling of cancer patient -derived cells (PDCs) and tissues. In addition to genetic profiling of original tumor tissues, the developed PDC models are characterized with genomics, image-based phenotyping (phenomics) and high throughput drug profiling. Here, we present how these methods are applied in our precision medicine project on solid tumors, specifically renal cancer. First, the original tumor tissue and the PDCs are analyzed by exome sequencing to ensure that somatic driver mutations and copy number variations are shared between the tissue and corresponding cell models. This is followed by phenomics approach, which allows comprehensive phenotypic characterization of tissues and ex vivo cell cultures, detection of biomarkers at very early stages, and functional analysis of drug responses at the single cell level. For the image-based drug profiling, the PDCs are treated with a library of >500 oncology compounds in five different concentrations, immunostained, and subjected to automated high-content imaging and image analysis. The features extracted from single cells are used for phenotypic classification based on machine learning to provide information on drug effects on the different cellular phenotypes. In our proof-of-concept study of clear cell renal carcinoma, PI3K/mTOR pathway inhibitors were found to be among the drugs inhibiting the proliferation of renal cancer PDCs, in agreement with detected somatic mutations affecting these pathways. In addition, our image-based drug sensitivity testing revealed the intra-sample heterogeneity in drug sensitivity and resistance. We are now further investigating the mechanisms of most potent drugs and their combinations in PDCs at single cell level. As a conclusion, our results implicate the importance of genomics and phenomics in comprehensive characterization of PDCs, and how phenomics enables the scoring of drug responses in heterogenic cancer cell cultures. We foresee that these approaches may potentially improve the translation of results back to clinic and support the design of combination therapies in cancer.

avatar for Davide Danovi

Davide Danovi

Director, HipSci Cell Phenotyping, King's College London
Dr. Davide Danovi is leading the HipSci Cell Phenotyping group at the Centre for Stem Cells and Regenerative Medicine at King's College London in the framework of the Wellcome Trust and MRC funded HipSci project. | | Davide holds an MD from the University of Milan and a PhD in ... Read More →

Emma Shanks

Head of Screening, Beatson Institute
The Screening Facility couples genetic and chemical high throughput screening with multiparametric phenotypic image analysis to support translational cancer research. We employ functional genomics (siRNA, shRNA, CRISPR) tools in both pooled and arrayed approaches to support targe... Read More →
avatar for Vilja	Pietiäinen

Vilja Pietiäinen

Senior Scientist, FIMM/University of Helsinki

Wednesday September 20, 2017 12:00pm - 2:30pm
Rosales I Courtyard by Marriot Madrid Princesa Hotel
Feedback form isn't open yet.