Seminars and Events

Past Events

Category Lecture
Date and Time 2017-02-21 11:00 - 12:30
Venue Auditorium C1F
Speaker Jan Ellenberg
Affiliation European Molecular Biology Laboratory (EMBL)
Title Systems biology by light and electron microscopy – from protein complexes via cellular protein networks to embryonic development
Poster click here to download(PDF)
Host Tomoya Kitajima
Summary New microscopy technologies such as super-resolving light and high-resolution electron microscopy are revolutionizing our ability to unravel structure and function of large molecular assemblies in situ. The nuclear pore complex (NPC) is the largest macromolecular protein complex in human cells and mediates all nucleocytoplasmic transport. Each NPC consists of several hundred proteins, and has the mass of many ribosomes. Although we understand the molecular composition and atomic structure of many parts of the NPC, how the whole complex assembles into the double membrane boundary of the intact nucleus has remained enigmatic. To capture assembly intermediates we correlated live cell imaging with high-resolution electron tomography and super-resolution microscopy. Surprisingly, assembly intermediates are dome-shaped evaginations of the inner nuclear membrane, which grow in diameter and depth until they fuse with the flat outer nuclear membrane. Super-resolution microscopy revealed the molecular maturation of assembly intermediates, which initially contain nuclear ring, and only later cytoplasmic filament proteins. NPC assembly thus proceeds by an asymmetric inside-out fusion of the inner with the outer nuclear membrane. Combining 3D super-resolution imaging with computational single particle averaging now allows us to unravel the molecular architecture of the NPC at nano-scale resolution.

High throughput microscopy allows us to study cellular functions in a systematic and quantitative manner. Cell division of somatic cells drives the continuous renewal of tissues and without control mechanism can lead to tumorigenesis. Using systematic gene silencing by RNAi and subsequent phenotyping by high-throughput live imaging we have defined ~600 genes that are required for a human cultured cell to divide normally. To understand the encoded protein network and its orchestration in space and time, we have established an integrated systems biology imaging workflow. After homozygous genome editing to functionally tag all endogenous copies of a given mitotic protein fluorescently, we image its absolute abundance and subcellular distribution by calibrated 4D imaging relative to spatio-temporal landmarks. Computational image analysis and modeling allows us to derive a canonical mitotic cell to integrate and navigate all protein data. Analysis of the protein fluxes between subcellular structures by machine learning can then predict dynamic protein clusters, the order of their formation and disassembly, and the abundance of their components. Our integrated computational and experimental method is generic and allows to map the dynamic protein networks that drive cellular processes.

Light-sheet imaging has revolutionized our understanding of cell division during embryonic development in many biological model systems. Early mammalian development, however, remained inaccessible due to high light sensitivity and demanding culture requirements. This is critical, since the embryonic aneuploidies that are a major cause of infertility and congenital diseases arise during the first divisions of the embryo. To address his, we have developed a novel inverted light-sheet microscope, which enabled the first in toto imaging of preimplantation mouse development from zygote to blastocyst. Using computational cell tracking, lineage tree reconstruction, and fate assignment revealed when fate specification happens. Furthermore, we could get the first insights about when and how chromosome segregation problems lead to aneuploidy and have discovered a surprising novel aspect of zygotic spindle formation that explains parental genome separation in early development.