Resolving Biostructures In-situ via Cryogenic Light and Electron Microscopy

Realizing the promise of cryogenic electron tomography (cryo-ET), alone or combined with cryogenic focused ion beam (cryo-FIB) milling, to deliver a proteomic atlas of cells requires further development in multiple areas. This project will integrate complementary expertise in cryogenic super-resolution fluorescence microscopy (cryo-SFM), high-resolution cryo-electron microscopy and correlative cryo-microscopy, artificial intelligence (AI)-based signal processing.

The team aims to incorporate an advanced fluorescence microscope inside a commercial cryo-FIB to improve the quality and efficiency of lamellae preparation. Unlike current combinations, this hybrid fluorescence microscope-cryo-FIB will incorporate point-spread-function engineering, providing unprecedented localization and identification accuracy (<100 nm) of subcellular features for 3D guidance of cryo-FIB milling. In addition, AI-based computational methods will help minimize noise and missing wedge artifacts in cryo-ET, which will allow for more accurate feature characterization of subcellular components such as membranes, organelles, and vesicles. The team will also develop an efficient pipeline that optimizes experimental and computational protocols in cryo-ET and subtomogram averaging, using T. gondii microtubules as a benchmark. This pipeline will enable researchers to determine structures of a major eukaryotic cytoskeletal component in situ. Achieving these goals will advance the construction of full proteomic atlases of cells in health and disease.