During my postdoctoral studies with Adam Frost at UCSF (San Francisco, USA), I addressed a long-standing question about how the nuclear envelope reforms at the end of mitosis. By bridging scales from single molecules to the cell biology level, I identified a new function of LEM2, a conserved protein embedded in the inner nuclear membrane, as a receptor for the membrane remodeling ESCRT-machinery. I characterized the mechanism that couples spindle microtubule disassembly to membrane fusion and discovered that LEM2 employs its intrinsically disordered domain to condense into a phospho-regulated fluid at the boundary between coalescing membrane sheets, the surface of the chromatin disk, and spindle microtubules during mitotic exit. In this arrangement, LEM2 activates and co-polymerizes with the ESCRT-III protein CHMP7. The model of LEM2’s architecture— a membrane protein condensate mediating spindle nuclear envelope interactions—is unprecedented in membrane remodeling biology and explains early nuclear compartmentalization, ESCRT-directed spindle disassembly, and eventual membrane sealing are coordinated. Both proteins, LEM2 and CHMP7, have recently been linked to human diseases, including progeria, cataract formation, sudden cardiac death, micronucleation, and cancer indicating the clinical relevance of this newly discovered pathway and mechanism.
During my PhD with Martin Beck at EMBL (Heidelberg, Germany), I studied the structure of the human nuclear pore complex. Nuclear pores direct the transport of biomolecules across the nuclear envelope and hence segregate transcription from translation. About 1000 protein building blocks assemble into a 110-MDa complex that fuses the inner and outer membranes of a cell's nucleus. Despite decades of research, the NPC’s sheer size and its inextricable position within the membrane of the nuclear envelope had prevented a detailed structural understanding. Efforts in x-ray crystallography resulted in a collection of highly resolved domains and fragments and it remained unclear how these fragments join to form the fully assembled structure seen in cells. The NPC thus remained a jigsaw puzzle of unknown shape consisting of pieces too small to sensibly put together. I developed integrated structural biology approaches to investigate the architecture of the human NPC. Combining mass spectrometry and Cryo-Electron tomography approaches allowed me to build the first structural models of the isolated human Y- and Nup214 complexes and to show how they assemble to form a functional NPC. I further developed and applied new modelling approaches to integrate the spatial restraints obtained by cross-linking mass spectrometry with the cryo-electron tomography data and other available high-resolution structures to overcome, piece by piece, the complexity of the NPC resulting in the first comprehensive model of this 110 megadalton membrane embedded complex covering ~2/3 of its components. This architectural model of the human NPC scaffold provided a first structural snapshot to understand the diverse aspects of nuclear transport, gene expression regulation, and how NPC components interplay with the nuclear envelope, creating a semipermeable organelle.
How biomolecules self-assemble functional membranous organelles depicts a blind spot in our current map of cellular organization. To address this, we study how the human cell’s biggest organelle, the nucleus, self-assembles following every cell division, very much like assembling a tent from its tightly packed pieces. In particular, we aim to understand the following aspects of this mystifying process: