Skip to end of metadata
Go to start of metadata

Visit the official Conference Site SIBS2010

Venue

EPFL (Ecole Polytechnique Fédérale de Lausanne), School of Life Sciences,  room SV1717A

Organizing and Scientific Committee

Chair: Gerardo Turcatti ; EPFL

Peter Horvath , Pauli Rämö , Berd Rinn , Andreas Vonderheit ; ETHZ
Ela Hunt , Peter Kunszt ; SyBIT

Program

Program in SIBS2010 website

Speakers

Gerlich Daniel Wolfram
daniel.gerlich@bc.biol.ethz.ch

Live-cell microscopy to screen for mitotic regulators

Fluorescence live cell microscopy is one of the most powerful tools to investigate complex cellular dynamics, for example in cell division, cell motility, or intracellular trafficking. The availability of large-scale RNAi technology and automated microscopy has opened the possibility to perform live cell imaging in functional genomics applications. This dramatically increases the content of a screening assay, yet it poses new challenges to achieve accurate quantitative annotation. We have developed a computational framework to annotate complex cellular dynamics and used this to screen by live imaging for phosphatases regulating mitotic exit in human cells.


Cameron Scott
cameron.scott@unige.ch

Dr Cameron Scott has a postdoctoral position in Jean Gruenberg lab at the University of Geneva. He screened at the EPFL screening platform (BSF) the siRNAs targeting the druggable genome in a highly informative lipids translocation assay.

High-throughput analysis of genes controlling
intracellular cholesterol transport

Cholesterol is both a major constituent of cellular membranes and an essential lipid critical for many metabolic and signaling pathways in the cell. Although mammalian cells are capable of the de novo synthesis of cholesterol, many, if not most cell types acquire the majority of their cholesterol exogenously through receptor-mediated endocytosis of cholesterol-carrying particles. Once internalized into endosomes the acidic milieu disengages cholesterol from their carriers, and through an ill-defined process, cholesterol is exported to the other organelles of the cell. Although the molecular details of cholesterol egress from the late endosome remain unresolved, this movement is dependent on at least one additional lipid, lysobisphosphatic acid (LBPA), and two disparate proteins, NPC1 and NPC2, defects in which cause the serious lysosomal-storage disorder Niemann-Pick disease, type C in humans. In order to identify novel genes involved in mediating endosomal cholesterol export, we have initiated microscopy-based, siRNA high-content screening methodologies to monitor lipid amounts and distributions within the cell. Using image analysis algorithms, we have simultaneously localized cholesterol and LBPA within the cell population after protein knock-down. Further, we have tested multiple cell-types to gain insight into the commonalities of intracellular cholesterol transport between divergent mammalian cells.


Anthony Nichols
anthony.nichols@merckserono.net

Primary, Secondary Screening and Genotoxicity Profiling of Compounds Using High-Content Imaging Platform.

The ever-pressing need to improve the yield of higher quality hits has led to the use of cell-based screening as a method of choice in drug discovery. High Content Screening (HCS) is multiplexed, functional cell-based screening. At Merck Serono Geneva Research Institute HCS cell-based assays are used in all aspects of drug discovery. The biological applications of HCS cell-based screening has been implemented, by using fluorescent plate imagers, in our research in signaling, GPCR, cell morphology and toxicology. HCS has enabled us to have an insight in the cellular effects of our clinical candidates in multiple cellular phenomena. HCS platforms have been placed within our core high throughput screening facility. Our protein and small molecule strategies converge on diseases in therapeutic areas such neurological disorders, autoimmune diseases and oncology. Our scientists use HCS for assay development, primary, secondary screening and toxicology testing.

Petr Strnad
petr.strnad@seznam.cz
(Previously EPFL, currently EMBL)

Petr Strnad finished his PhD thesis in Prof. Pierre Gönczy Lab at the EPFL Cancer Institute (ISREC). In collaboration with the EPFL screening platform (BSF), he achieved the first high content RNAi whole genome screening

Genome-wide Image-based siRNA Screen for Genes Involved in Centrosome Duplication

During mitosis, a bipolar microtubule-based structure called the mitotic spindle faithfully segregates chromosomes into two daughter cells. The centrosome is the main microtubule organizing center of animal cells; there are two centrosomes present in mitosis that act as the two spindle poles, thus contributing to spindle bipolarity. Cells have only a single centrosome early in the cell cycle, which must duplicate exactly once to give rise to the two centrosomes acting during mitosis. The mechanisms governing centrosome duplication are only partially understood.
To better understand the molecular basis of centrosome duplication in human cells, we conducted a genome-wide image-based screen to identify genes required for this process. Human cells harboring GFP-labeled centrosomes and treated by siRNAs were imaged and the number of centrosomes in each cell automatically counted using a custom algorythms developped in Matlab and KNIME. This provides us with a direct readout of centrosome number and thus also successful centrosome duplication.

Lucas Pelkmans
lucas.pelkmans@ethz.ch

Cell-to-cell variability modelling reveals complex RNAi phenotypes in virus infection

Isogenic mammalian cells display regulated cell-to-cell variability established during the growth of a population. To advance quantitative and predictive approaches in mammalian cell biology, this fundamental variability must be incorporated in our interpretation of cell population-wide perturbation effects. I will present how we achieve this for the study of virus infection. I will present parallel RNA interference screens of 17 different mammalian virus infections in up to 4 different human cell lines per virus, which allowed us to arrive at global statements. I will present a comprehensive computational approach that employs Bayesian and novel multivariate methods to model the effects of each perturbation at the single-cell level. This approach reveals the complex causal interactions that lead to RNAi phenotypes of virus infection in a population of cells, which is essential for a meaningful classification of viruses based on their host factor dependencies. Finally I show that this analysis enabled the discovery of a new class of phenotype that acts on the cell-to-cell variability pattern of infection.

Roger Meier
roger.meier@ethz.ch

Roger Meier is a PhD student in Prof. Ari Helenius Lab at the Institute of Biochemistry at the ETH Zürich. He performed a whole genome screen together with the RISC facility at ETH

siRNA Infection Screens: Pitfalls and Promise

In recent years siRNA silencing has emerged as a powerful tool in infection biology especially when combined with automated high-throughput infection screening. To study bunyavirus entry into HeLa cells, we+ used Uukuniemi virus (UUKV) in a genome wide siRNA-silencing screen. Cells were transfected by different siRNAs against each gene in 384-well plates, and infected with UUKV 72 h later. The infected cells were identified by immunostaining against the viral nucleoprotein N using automated fluorescence microscopy. Infection was quantified by computational image-based analysis using «CellProfiler» and «Advanced Cell Classifier». A ranked list of potential hits involved in UUKV infection was determined according to a redundant siRNA activity probability-based approach. To identify clusters of predicted functional protein interactions DAVID and STRING database were used. While siRNA infection screening has great potential, as a powerful new approach to analyze pathogen/host interactions, there are many technical difficulties and pitfalls, starting with siRNA design and ending with data analysis. We will discuss these challenges based on our UUKV-screen, and suggest approaches how to overcome some of them.

Marjo Götte
marjo.goette@novartis.com

High-content screening at the Lead Finding Platform of Novartis Institutes of BioMedical Research

High-content screening (HCS) has become an important and established technology in various phases of drug discovery. The technology, including reagents, imagers, image and data analysis, is still rapidly developing, and it enables performance of assays in high throughput that earlier were possible only in small scale. At Novartis Institutes of BioMedical Research high-content screening is intensively used in identification of drug targets and drug candidates, and in profiling of lead compounds. HCS provides more information about compounds at an early phase than traditional biochemical or cellular screening. In the presentation I will tell about high-content screening at Novartis having emphasis on HCS at the Lead Finding Platform. Examples of different high-content screening projects will be provided.

Michael Prummer,
Discovery Technologies, F. Hoffmann - La Roche Ltd
michael.prummer@roche.com

Automation of early genotoxicity predictions in non-adherent cells by image-based screening

Most High-Content Screening assays are run in microtiter plate format with adherent cells and not with suspension cells due to practical limitations. However, there are established cellular models for conventional microscopy in which suspension cells are used. We are presenting here a novel and universal approach to fabricate arrays of non-adherent cells for automated microscopy that fills this gap. As a first application we report on a high-content toxicity assay at the single-cell level using an innovative new preparation tool. For image segmentation and object classification we have co-developed an adaptive and iterative algorithm using Definiens' Cognition Network Technology. From the derived data we extract a toxicity score by non-parametric statistical methods for fully automated compound classification. Taken together, these developments open up new routes for miniaturized high-content screening with suspension cells.

Thomas Wild
e-mai


Thomas Wild is a PhD student in Prof Ulrike Kutay's Lab at the Institute of Biochemistry at ETH Zürich and performed the first whole genome screen in Switzerland together with the RISC facility

An RNAi Screening Approach to Identify Human Ribosome Synthesis Factors

The synthesis of ribosomal subunits is a complex, multistep process so far studied mostly in unicellular organisms. In yeast, more than 200 factors are required for the assembly of 40S and 60S ribosomal subunits. To systematically identify proteins involved in ribosome biogenesis in human cells, we used an RNA interference (RNAi) approach. In a first step, we analyzed the contribution of 464 candidate factors to human ribosome synthesis. Using inducible, fluorescent ribosomal proteins as reporters, we visually detected defects both in 40S and 60S biogenesis. By performing computer-based image analysis based on supervised machine-learning techniques we obtained evidence for a functional link of 153 human proteins to ribosome synthesis. This dataset shows that core features of ribosome assembly are conserved from yeast to human, however, differences exist for instance with respect to 60S subunit export. Aiming towards a comprehensive list of human ribosome synthesis factors, we next performed a genome-wide RNAi screen scoring for defects in nuclear 40S biogenesis. Analysis of primary screening results revealed not only consistency with the outcome of  the candidate screen, but also that many novel 40S biogenesis factors have likely been identified. We are currently setting up a secondary screen to verify the results from the genome-wide screen.

Sebastian Maerkl
sebastian.maerkl@epfl.ch


A high-throughput microfluidic platform for single cell screening

We have developed a massively parallel microchemostat platform capable of growing 1152 bacterial or yeast strains in parallel. The platform is integrated to allow automated image acquisition with high spatio-temporal resolution. With this platform we conducted the first proteome wide survey of protein expression dynamics in S. cerevisiae in response to MMS induced DNA damage. We also applied the platform to screening of a M. smegmatis knock-out library (the model organism for M. tuberculosis) for mutants which are hyper- or hypo -sensitive to an antibiotic yielding potential drug targets and basic insights into bacterial persistence.

Ray Jones
thouis@gmail.com

Approaches to analyzing individual cell data in high content screens

High-content screening allows access to vast amounts of data on
individual cells. Analyzing this data on a per-cell level can be very
powerful, particularly when a screen's readouts are complex or rare
phenotypes, but dealing with such large data sets can also be
challenging. This talk will discuss methods we have developed for
analyzing per-cell data, including "virtual flow cytometry,"
interactive phenotype classification, and large-scale machine
learning, along with more recent work on automated methods for
phenotype discovery.

 

Eugenio Fava
e-mail

title



Christian Conrad
cconrad@embl.de

Automation of f-techniques in hcs-screening

Quantitative confocal fluorescence microscopy in cell biology, including advanced F-techniques (FRAP, FLIM, and FCS), relies on higher sample numbers to accomplish robust statistics. It becomes increasingly more important to acquire massively single cell high-resolution image data in an unbiased manner. To efficiently identify such distinct cellular subpopulations or cell changes we engineered an 'Micropilot' confocal microscope software suite facilitating supervised machine learning and the interface communication with various microscope systems during screening process. Such automated platforms clearly provide a broad range of applications in cell and system biology.

 

Peder Zipperlen
peder.zipperlen@gmail.com

Experiences with setting up and operating an
automated high-throughput image-based RNAi
screening platform in a biotech environment

Recent years have seen a significant increase in the development of automated systems for cell based screening applications. The requirements on such platforms are significantly higher than for standard lab automation systems. Even more so when operated under biosafety level 2 conditions. Based on a system used at 3-V Biosciences for both, large scale RNAi and compound screening, design requirements and operational challenges will be discussed. A special emphasis will be given on screen design as well as RNAi and compound logistics.

Matthias Lutolf
matthias.lutolf@epfl.ch
Research in Prof. Lutolf lab is at the interface of stem cell biology and biomolecular engineering to gain fundamental insight into how protein components of tissue-specific microenvironments, termed niches, control the behavior of multipotent adult stem cells such as hematopoietic stem cells (i.e. the stem cells that give rise to all blood cells).

High-throughput methods to define complex stem cell niches

The potential of stem cells in clinics and as a diagnostic tool is still largely unmet, partially due to a lack of in vitro models that efficiently mimic the in vivo stem cell microenvironment-or niche-and thus would allow reproducible propagation of stem cells or their controlled differentiation in vitro. The current methodological challenges in studying and manipulating stem cells have spurred intense development and application of microfabrication and micropatterning technologies in stem cell biology. I will discuss selected approaches that my lab has developed to dissect the complex molecular interplay of stem cells and their niche and study single-cell behavior in high-throughput. Increased merging of microfabrication with advanced biomaterials technologies may ultimately result in functional artificial niches capable of recapitulating extrinsic stem cell regulation in vitro and on a single-cell level.

M. Podvinec
michael.podvinec@unibas.ch

iBRAIN2 - reengineering an image analysis workflow system

Our new software, a rewrite of iBRAIN previously developed at the ETH Zurich, supports automated image analysis for SystemsX.ch projects at several locations and uses various computer cluster architectures. We now have a working system at Basel which allows the users to automatically perform data analysis and are working on a number of system enhancements and porting it to other institutions. Current work focuses on the integration with openBIS and B-fabric databases and with the Cell Profiler software from the Broad Institute. I will discuss the achievements and technical challenges facing us and our strategy to solve those. The work is part of SyBIT, an IT project supporting a wide range of computational needs in several systems biology projects funded by SystemsX.ch.


Pauli Rämö
raemoe@imsb.biol.ethz.ch

Statistical Analysis of High-Content siRNA Screens

High-content siRNA screening has revolutionized functional screening in mammalian cells. High-content screening involves many wet-lab and computational steps, each of them being a possible source or errors that may invalidate the final conclusions of the data. We propose a general and complete high-content analysis pipeline and validate each step against other methods published in the literature. By comprehensive quantification we discover which steps are the most important and propose screening design principles that will yield the best possible results. We find our that comprehensive and careful computational analysis is not only crucial for correct data interpretation, but can alleviate some problems occurred in the web-lab stage. We also propose new methodology to integrate interaction networks with screening data.

  • No labels