A workflow for 3D-CLEM investigating liver tissue

A Kremer,E VAN Hamme,J Bonnardel,P Borghgraef,C J GuÉrin,M Guilliams,S Lippens

DOI: https://doi.org/10.1111/jmi.12967

Abstract:Correlative light and electron microscopy (CLEM) is a method used to investigate the exact same region in both light and electron microscopy (EM) in order to add ultrastructural information to a light microscopic (usually fluorescent) signal. Workflows combining optical or fluorescent data with electron microscopic images are complex, hence there is a need to communicate detailed protocols and share tips & tricks for successful application of these methods. With the development of volume-EM techniques such as serial blockface scanning electron microscopy (SBF-SEM) and Focussed Ion Beam-SEM, correlation in three dimensions has become more efficient. Volume electron microscopy allows automated acquisition of serial section imaging data that can be reconstructed in three dimensions (3D) to provide a detailed, geometrically accurate view of cellular ultrastructure. In addition, combining volume-EM with high-resolution light microscopy (LM) techniques decreases the resolution gap between LM and EM, making retracing of a region of interest and eventual overlays more straightforward. Here, we present a workflow for 3D CLEM on mouse liver, combining high-resolution confocal microscopy with SBF-SEM. In this workflow, we have made use of two types of landmarks: (1) near infrared laser branding marks to find back the region imaged in LM in the electron microscope and (2) landmarks present in the tissue but independent of the cell or structure of interest to make overlay images of LM and EM data. Using this approach, we were able to make accurate 3D-CLEM overlays of liver tissue and correlate the fluorescent signal to the ultrastructural detail provided by the electron microscope. This workflow can be adapted for other dense cellular tissues and thus act as a guide for other three-dimensional correlative studies. LAY DESCRIPTION: As cells and tissues exist in three dimensions, microscopy techniques have been developed to image samples, in 3D, at the highest possible detail. In light microscopy, fluorescent probes are used to identify specific proteins or structures either in live samples, (providing dynamic information), or in fixed slices of tissue. A disadvantage of fluorescence microscopy is that only the labeled proteins/structures are visible, while their cellular context remains hidden. Electron microscopy is able to image biological samples at high resolution and has the advantage that all structures in the tissue are visible at nanometer (10-9 m) resolution. Disadvantages of this technique are that it is more difficult to label a single structure and that the samples must be imaged under high vacuum, so biological samples need to be fixed and embedded in a plastic resin to stay as close to their natural state as possible inside the microscope. Correlative Light and Electron Microscopy aims to combine the advantages of both light and electron microscopy on the same sample. This results in datasets where fluorescent labels can be combined with the high-resolution contextual information provided by the electron microscope. In this study we present a workflow to guide a tissue sample from the light microscope to the electron microscope and image the ultra-structure of a specific cell type in the liver. In particular we focus on the incorporation of fiducial markers during the sample preparation to help navigate through the tissue in 3D in both microscopes. One sample is followed throughout the workflow to visualize the important steps in the process, showing the final result; a dataset combining fluorescent labels with ultra-structural detail.