Anatomy measurements in the brain are a serious challenge, combining a need for high resolution and large volumes: neural processes and synapses are on the order of 30 nm, but branches may reach across the entire brain – hundreds of microns in the fly and millimeters in the mouse. Electron microscopy is the gold standard for achieving the necessary resolution, but large volumes and specific labeling of neurons of interest are difficult. Fluorescence microscopy allows specific labeling of neurons and proteins and imaging over large volumes, but conventional techniques do not have adequate resolution. We have two projects that help bridge that gap:
We have built a high throughput array tomography microscope, which achieves high resolution through imaging of ultrathin slices (read about some current applications in the Spruston lab here). The acquisition of images of brain slices, whether taken by optical or electron microscopy, presents a formidable downstream challenge to first align and then trace neurons. We have developed and made available a fast, high-capacity serial-section aligner program to address the alignment issue.
We are currently developing a Bessel-beam selective-plane structured illumination microscope for imaging cleared fly brains at high resolution. This instrument accommodates whole brain samples and achieves resolution of 170x170x300 nm. Although that still falls short of EM resolution, it is fully 3X better than conventional confocal microscopy.