our multidisciplinary approach
The Gerlich lab brings together a diverse group of scientists with backgrounds in cell biology, biochemistry, physics, and computer science to investigate chromosome organization in creative ways. Because we like to ask unconventional questions, we often must establish new assays to visualize cellular processes. We exploit a broad panel of advanced microscopy techniques and combine them into correlative workflows to bridge the gap in resolution between molecular and cell biology. This collaborative approach has allowed us to discover novel biological phenomena for which molecular regulators have not yet been described.
To elucidate the molecular foundation of these previously undescribed phenomena, we perform image-based loss-of-function screening. Computer scientists in our laboratory have implemented CellCognition: a fully automated, live-cell microscopy platform to perform cell phenotype analysis by machine learning (see below). New molecular regulators identified by image-based screening are systematically characterized in cells using state-of-the-art molecular techniques, including CRISPR/Cas9-based genome engineering, functional readouts using biosensors, chemically-induced protein dimerization, and micro-manipulation using intracellular laser surgery.
Once we have determined the cellular functions of novel regulators, we aim to dissect the underlying molecular mechanisms. We characterize molecular properties using a variety of biochemical technologies, including systematic analysis of molecular interactions and posttranslational modifications by mass spectrometry.
We then use in vitro reconstitution with purified components to test hypotheses about molecular mechanisms. For example, biochemists in our team have established novel assays to study the mechanical properties of mitotic chromosomes using purified chromatin, DNA-coated beads, and recombinant proteins. We manipulate the components in this system using microfluidic chambers and magnetic force control, and analyze biophysical properties using light- and atomic-force microscopy. Ultimately, our goal is to reconstitute molecular activities using synthetic components that mimic the functions of native cellular proteins.
The ability to pursue an idea that begins with a general description of cell biological process and results in a detailed understanding of mechanistic details of the molecular components requires a broad range of technologies. We are able to accomplish this because of the outstanding research infrastructure available at campus facilities, and our strong network of international collaboration partners. Researchers in our laboratory take the lead in developing hypotheses and establishing the experimental assays and approaches to address them, but they are extensively supported by our facility staff. This teamwork approach leverages our research beyond the boundaries of individual biological disciplines, and allows us to accomplish ground-breaking science.
Technologies and model systems
The cutting-edge technologies provided by 18 scientific facilities at our campus offer an unmatched opportunity to focus entirely on biological questions, rather than troubleshooting technical details. At IMBA, we have free access to a variety of microscopes at the forefront of imaging through the BioOptics, VBCF Advanced Microscopy, and Electron Microscopy facilities. This large collection of scanning- and spinning disc confocal microscopes, automated screening microscopes, laser microsurgery, super-resolution fluorescence, and electron microscopes enables us to approach scientific questions from multiple angles.
Our biochemical experiments are supported by the Protein Chemistry, Protein Technology, and Metabolomics facilities, and chromosome conformation capture experiments are carried out in collaboration with the Next Generation Sequencing facility. Genome engineering and molecular biology services are further supported by Protein Technology and Molecular Biology facilities.
We use human tissue culture cells as a main model system and have developed a large library of cell lines which stably express various combinations of fluorescent marker proteins for systematic phenotype profiling. Our experiments are typically highly automated, involving high-throughput transfection and automated microscopy for high-content screening. To efficiently extract quantitative phenotype traits from these large-scale microscopy experiments, we have developed computer vision and machine learning methods that are disseminated as the open-source software CellCognition.
CellCognition and CellCognition Explorer are open source software programs for the analysis of cellular phenotypes. Developed in our laboratory, they combine object detection, single-cell tracking, feature extraction, and classification of morphologies (Figure 1) for automated annotation of different cell morphologies or kinetic readouts of cell features. CellCognition has been optimized for large-scale screening data, runs on Windows, Mac OS X, and Linux desktop computers, and provides an interface to computing clusters for ultra-high-throughput data analysis (see Held et al., Nature Methods 2010). CellCognition Explorer provides an advanced user interface for interactive classifier training of medium-scale data, and further provides novelty detection methodology (see Sommer, Hoefler et al., MBoC 2017).
References for computer vision and machine learning methodology
C. Sommer, R. Hoefler, M. Samwer, D.W. Gerlich. A deep learning and novelty detection framework for rapid phenotyping in high-content screening. Mol Biol Cell. (2017): 28(23): 3428-3436. doi: 10.1091/mbc.E17-05-0333
Sommer C, Held M, Fischer B, Huber W, Gerlich DW. (2013). CellH5: a format for data exchange in high-content screening. Bioinformatics. 29(12):1580-2
Zhong Q, Busetto AG, Fededa JP, Buhmann JM, Gerlich DW. (2012). Unsupervised modeling of cell morphology dynamics for time-lapse microscopy. Nat Methods. 9(7):711-3
Held M, Schmitz MH, Fischer B, Walter T, Neumann B, Olma MH, Peter M, Ellenberg J, Gerlich DW. (2010). CellCognition: time-resolved phenotype annotation in high-throughput live cell imaging. Nat Methods. 7(9):747-54
Sommer C, Gerlich DW. (2013). Machine learning in cell biology - teaching computers to recognize phenotypes. J Cell Sci. 126(Pt 24):5529-39
Conrad C, Gerlich DW. (2010). Automated microscopy for high-content RNAi screening. J Cell Biol. 188(4):453-61.
References for screening applications
Samwer, M., Schneider, MWG., Hoefler, R., Schmalhorst, PS., Jude, JG., Zuber, J., Gerlich, DW. (2017). DNA cross-bridging shapes a single nucleus from a set of mitotic chromosomes. Cell 170(5):956-972.e23
Cuylen, S., Blaukopf, C., Politi, AZ., Müller-Reichert, T., Neumann, B., Poser, I., Ellenberg, J., Hyman, AA., Gerlich, DW. (2016). Ki-67 acts as a biological surfactant to disperse mitotic chromosomes. Nature. 535(7611):308-12
Wurzenberger C, Held M, Lampson MA, Poser I, Hyman AA, Gerlich DW. (2012). Sds22 and Repo-Man stabilize chromosome segregation by counteracting Aurora B on anaphase kinetochores. J Cell Biol. 198(2):173-83
Piwko W, Olma MH, Held M, Bianco JN, Pedrioli PG, Hofmann K, Pasero P, Gerlich DW, Peter M. (2010). RNAi-based screening identifies the Mms22L-Nfkbil2 complex as a novel regulator of DNA replication in human cells. EMBO J. 29(24):4210-22.
Schmitz MH, Held M, Janssens V, Hutchins JR, Hudecz O, Ivanova E, Goris J, Trinkle-Mulcahy L, Lamond AI, Poser I, Hyman AA, Mechtler K, Peters JM, Gerlich DW. (2010). Live-cell imaging RNAi screen identifies PP2A-B55alpha and importin-beta1 as key mitotic exit regulators in human cells. Nat Cell Biol. 12(9):886-93.