Metastatic events are transient and rare, making them difficult to observe in vivo. Intravital microscopy (IVM) enables studing crucial steps in dissemination of tumor cells in living organisms, but is limited in resolution. Combining IVM to 3D Electron Microscopy (3DEM) enables to correlate functional and dynamic in vivo imaging to high-resolution of the tumor cells and their microenvironment. However, keeping track of single tumor cells when moving from IVM to EM imaging is highly challenging in complex voluminous samples.
Here, I will present a method that exploits x-ray microscopic computer tomography (microCT) to correlate IVM to EM. First, fluorescent tumor cells are xenotransplanted to mouse brain and skin tissue, and monitored by IVM. 3D datasets of the tumor cells and the surrounding fluorescently stained vasculature are obtained. Next, the region of interest is processed for EM. The resin-embedded sample is then scanned with microCT, revealing the outlines of the resin block, the biopsy and the vasculature within. Using 3D imaging software, the 3D IVM and microCT datasets are registered, enabling to map the position of the tumor cell within the resin-embedded specimen. Finally, targeted trimming enables to quickly approach the tumor cell inside the resin block and expose it for 3D imaging with FIB-SEM or serial electron tomography.
The method will be demonstrated on capturing arrested tumor cells in the vasculature of mouse brain samples, and on migrating invasive tumor cells in mouse skin. Enabling to predict the position of the tumor cell within the resin block with an accuracy of >5 µm significantly speeds up the process of correlating IVM to EM; from several months to ~2 weeks. Facilitating the correlative approach uniquely enables multiple high-resolution observations of rare metastatic events in tissue, allowing obtaining statistically relevant conclusions on the crucial steps in the dissemination of tumor cells.