A workflow for streamlined acquisition and correlation of serial regions of interest in array tomography

Background
Array tomography (AT) is a high-resolution imaging method to resolve fine details at the organelle level and has the advantage that it can provide 3D volumes to show the tissue context. AT can be carried out in a correlative way, combing light and electron microscopy (LM, EM) techniques. However, the correlation between modalities can be a challenge and delineating specific regions of interest in consecutive sections can be time-consuming. Integrated Light and Electron microscopes (iLEMs) offer the possibility to provide well-correlated images and may pose an ideal solution for correlative AT. Here we report a workflow to automate navigation between regions of interest.
Results
We use a targeted approach that allows imaging specific tissue features, like organelles, cell processes, nuclei etc. at different scales to enable fast, directly correlated in situ AT using an integrated light and electron microscope (iLEM-AT). Our workflow is based on the detection of section boundaries on an initial transmitted light acquisition that serves as a reference space to compensate for changes in shape between sections, and we apply a stepwise refinement of localizations as the magnification increases from LM to EM. With minimal user interaction, this enables autonomous and speedy acquisition of regions containing cells and cellular organelles of interest correlated across different magnifications for LM and EM modalities, providing a more efficient way to obtain 3D images. We provide a proof of concept of our approach and the developed software tools using both Golgi neuronal impregnation staining and fluorescently labeled protein condensates in cells.
Conclusions
Our method facilitates tracing and reconstructing cellular structures over multiple sections, is targeted at high resolution
ILEMs, and can be integrated into existing devices, both commercial or custom-built systems.