From 2008 to 2013 we supported this consortium, which used hardware, software and biological tools separately developed in laboratories at three US institutes to collectively enable the construction of “connectomes” – wiring diagrams of whole regions of the nervous system.
The brain operates largely by transferring information between neurons at points of contact called synapses. Despite elegant anatomical studies over the last century, an actual wiring diagram that maps these synaptic connections has not yet been obtained for any animal except the roundworm nematode C. elegans.
Many neuroscientists are convinced that such a map for a higher level organism, such as a mouse, is critical to further studies of the brain. The hope is that the data will both guide and constrain mechanistic explanations of cognitive processes, and provide the foundation for determining how neural circuitry evolves and changes under physiological and pathological conditions. Recent technical innovations have led many to believe that constructing such a map may now be possible, and the Connectomics Consortium contributed to this effort through a range of new tools developed at separate laboratories across the United States of America.
The Consortium included the laboratories of Professors Jeff Lichtman, Josh Sanes and Xiaowei Zhuang at Harvard University, Professor Stephen Smith at Stanford University and Professor Sebastian Seung at the Massachusetts Institute of Technology.
Scientific approaches in the Consortium include reconstructions at a very high resolution using light and electron microscopy. The work developed novel protocols for staining and innovative techniques for improving the speed of imaging, automated lathe processing and high throughput array tomography. New antibodies have been developed to mark synaptic types for classification using the latter.
The Consortium improved the ability to superimpose light microscopy data on electron microscopy images of the same sample. Additional technology in use was the STORM microscope, which uses light rather than electrons to trace neural connections in labelled tissue samples. Involving the clever use of fluorescent probes, this approach enables the visualisation of fine grade neural ‘wires’ and synapses that are smaller than the diffraction limit of visible light. New mouse lines have been created with more coverage, higher levels per neuron and a greater colour range for use on light microscopy platforms.
Computational methods were developed to reassemble the serial sections into three-dimensional images and to trace connections automatically. Collaboration between the groups has shown that only minor alignment is necessary to get perfect registration of serial sections. This has somewhat superseded the previously important (yet tedious) work of human tracing by hand!