The goal is to create the largest ever roadmap of connections in the mammalian brain. Our electron microscopy team is contributing to this enormous effort by imaging billions of tiny synaptic connections in a cubic millimeter of mouse neocortex. “What we’re trying to do here is understand the brain on its own terms,” explains Clay Reid, Senior Investigator.

Within a cubic millimeter of brain, about the size of a grain of sand, there are a hundred thousand neurons and nearly a billion synapses. In order to see those synapses, the team performs a skillfully refined and challenging process with little room for error. Imaging these connections involves sectioning the brain sample into 25,000 pristine slices that are 40 nanometers wide, a fifth of the thickness of a human hair, then distributing these 25,000 slices over six electron microscopes to be imaged for four to six months. They then reassemble this volume and deliver them to collaborators at Princeton University.

When they receive the images, researchers at Princeton begin segmenting each one of the 100,000 neurons and finding all the connections between them. They then deliver back a wiring diagram and the morphology of these neurons to the Allen Institute  for the collaborative team to analyze together. “When we first saw the images come from Princeton after they finished their segmentations and we got to see what the data looked like, it was incredible,” explains Shelby Suckow, Project and Alliance Manager.

More than a decade ago, we got the first glimmer that we may be able to do something like this. And over the past year and a half, we built a pipeline that allows us to collect petabytes of data in a short period of time, and by working closely with our collaborators we have also optimized the segment reconstruction process. “In a year from now we expect to have, for the first time, a volume that will allow us to look at this circuit in the mouse,” explains Nuno Maçarico da Costa, Assistant Investigator. At the end of this process we will have wonderful three-dimensional images of neurons with the completeness and complexity of which no one has ever seen.