Brain vasculature imaging with two-photon and light-sheet microscopy

Cerebral blood vessels are charged with the fundamental task of ensuring and regulating blood supply depending on neuronal demand, through a mechanism known as “neurovascular coupling”. However, despite its essential role, we still lack a complete map of brain vasculature network on a brain wide scale. A fine morphological analysis of the brain vasculature requires a clear visualization of the finest capillaries. The resolution needed for visualizing the capillary network is achievable with optical microscopy. Nevertheless, light scattering limits the applications of optical microscopy for imaging inside intact tissues. To access deep regions, serial sectioning methodologies can be applied. These approaches, however, require long time for imaging large volumes, such as a whole mouse brain. Second, the deformations due to the cutting procedure make difficult a precise alignment of the images for a volumetric rendering. Finally, samples are destroyed during imaging, and are not available for subsequent investigation. Another approach consists on the application of “clearing methods” to render whole rodent brains transparent and so suitable for optical investigation without sectioning. Because of the high image acquisition speed, light-sheet fluorescent microscopy (LSFM) have been used in combination with clearing methods for whole brain investigation. However, vasculature staining used so far with clearing methods relies on lectins, which only label the blood vessels wall. This labeling offer a poor image contrast and makes arduous a proper vasculature reconstruction with automatic softwares. Indeed, no software based vascular network reconstruction in the whole brain has been performed. In the present thesis, a methodology for ex vivo analysis of whole mouse brain vasculature with micrometric resolution is presented. It consist of the use of LSFM in combination with a clearing method and a specialized blood vessels labeling able to fill the blood vessels lumen. The procedure preserve endogenous fluorescence, allowing for simultaneous or consecutive imaging of vessels and neurons in transgenic animals expressing fluorescent proteins. The high image contrast, along with vessels lumen labeling, allows for a better application of automatic tools for image segmentation, by means blood vessels are distinguished from the surrounding space. This represent a key step for a reliable tracing of the vascular network in a whole brain scale. Possible application have been evaluated using two-photon fluorescent microscopy (TPFM) on portions of mouse brain cortex using a mouse model of cortical stroke. These inspections reveal the potentiality of the combination between vascular lumen staining and tissue clearing for analysis of morphological changes of blood vessels. In addition, two strategies are explored for a clear distinction of the arterial and venular component. The establishment of a reliable differential staining for arteries and veins would permit an easier comprehension of the arterial and venular routes inside the brain. A detailed characterization of the brain vascular network will have a positive impact also for a correct interpretation of functional imaging techniques, such as BOLD-fMRI, which reveal changes on the oxygenation level in venous blood as endogenous contrast for visualizing brain areas with increased cerebral activity. Noteworthy, the fast image acquisition of the method proposed will foster a proper