NSF/BIO-UKRI/BBSRC: Comparing ultrastructure measures across multiscale electron microscopies

Imaging biological structures across scales, from individual molecules up to complex organs is possible using electron microscopy (EM). Two EM technologies have advanced rapidly in the past decade: 1) cryo-electron tomography (cryoET), which resolves molecules inside cells at near-atomic detail, and 2) volume electron microscopy (VEM), which maps cellular architecture of tissues across millimeter scales. Used together, they could reveal how molecules are organized from cells to tissues to support essential processes and functions. However, cryoET and VEM have evolved within separate research communities supported by incompatible software and data standards. The absence of a shared framework prevents connecting observations across different scales, leaving the field unable to construct integrated, multiscale models of tissue organization. This project will develop the imaging datasets, computational tools, standardized analysis pipelines, and unified metadata framework needed to close this gap. The approach combines artificial intelligence (AI)-enabled segmentation and classification with state-of-the-art cryoET and VEM imaging of tissues of organs central to neurological, metabolic, and cardiovascular health. The project will also produce the first publicly available multiscale 3D reference image libraries for these tissues, establishing open benchmark resources for the global research community. This work will build partnerships that unite US investigators at the University of Pittsburgh with UK partners at the Science and Technology Facilities Council. Together, the findings will accelerate biological discovery; train a new generation of researchers in skills at the interface of biology, AI, and data science; and strengthen research infrastructure, contributing to American leadership in biotechnology. This research project will establish a quantitative computational framework linking ultrastructural measurements made by cryoET, serial cryo-volume EM (cryo-VEM), and room-temperature volume EM (rtVEM) of the same tissue types. Although the field has cataloged ultrastructural features at each individual scale, the correspondences between molecular-resolution and tissue-scale measurements remain largely unexplored. Four coordinated aims address this gap. Aim 1 will generate systematic multiscale 3D reference image libraries from brain, pancreas, and heart tissue, with paired stimulated and unstimulated conditions. Aim 2 will develop cryoET analysis pipelines for automated segmentation, particle detection, and geometric quantification within the established Collaborative Computational Project for Electron cryo-Microscopy (CCP-EM) framework. This will include adoption of AI-based classification and structured metadata output. Aim 3 will develop VEM analysis pipelines for consensus segmentation, feature classification, and morphometric quantification, coordinating with the Collaborative Computational Project for Volume EM (CCP-volumeEM). Aim 4 will integrate cryoET and VEM measurements for cross-scale comparisons and use this correspondence to build multiscale models that incorporate molecular-resolution detail within tissue-scale anatomical contexts. Expected outcomes include validated multiscale benchmark datasets, AI-driven analytical tools released openly through CCP-EM, and development of a potentially transformative foundation for novel multiscale structural biology investigations that integrate molecular organization and tissue-level architecture. By developing this infrastructure for the biological sciences community, the project advances NSF priorities in Biotechnology and Artificial Intelligence. This award is made possible through the NSF-UKRI lead agency opportunity. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2624282 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Zachary Freyberg | Institution: University of Pittsburgh, PITTSBURGH, PA | Award Amount: $1,065,000 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2624282 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2624282.html