CMIC Seminar: Correlative Light and Electron Microscopy Goes Large: Dealing with the data deluge at the nanoscale

Speaker: Lucy Collinson PhD, Head of Electron Microscopy, London Research Institute, Cancer Research UK
UCL Contact: Dominique Drai (Visitors from outside UCL please email in advance).
Date/Time: 23 Jul 14, 12:00 - 13:00
Venue: Roberts 1.06
Further Information:

Light refreshments will be served

Abstract

Fluorescence microscopy is a powerful tool for localising proteins within biological samples. However, information is limited to the distribution of the tagged protein, telling us little about the ultrastructure of the surrounding cells and tissues, which may be intimately involved in the biological process under study. Electron microscopy overcomes the resolution limitation inherent in light microscopy and can reveal the ultrastructure of cells and tissues. However, protein localisation tends to be complex. Correlative light and electron microscopy (CLEM) combines the benefits of fluorescence and electron imaging, revealing protein localisation against the backdrop of cell and tissue architecture.

In this talk, I will introduce several ways in which we are moving towards 3D CLEM of whole cells and tissues. We use three different automated data collection systems to acquire huge 3D datasets (terabytes per week) detailing cell and tissue ultrastructure at nanometer resolution (Focused Ion Beam Scanning Electron Microscopy, Serial Block Face Scanning Electron Microscopy and Array Tomography). In the near future, massively-parallel imaging in electron microscopes means that data acquisition rates are likely to reach terabytes per hour. To this, we add 3D functional information through correlation with fluorophore-labelled proteins.

We now find that the major barrier to discovery in biological imaging is our ability to interrogate the data. This is partly due to a lack of algorithms for tracking regions of interest during data collection, a lack of segmentation algorithms to semi-automatically select structures of interest within cells and tissues, and a lack of algorithms for overlay of multi-modality imaging data to identify the subcellular location of proteins involved in disease states. As with all challenges, the flip-side is the potential for game-changing advances, in this case with the aim of linking medical imaging with molecular imaging for discovery research and personalised medicine.