Investigator: Edward Boyden Biological Engineering and Brain and Cognitive Sciences, MIT
Dr. Boyden, a Professor in Neurotechnology at MIT and Howard Hughes Medical Institute (HHMI) investigator, leads the Synthetic Neurobiology Group that has developed many cutting-edge techniques for understanding brain connectivity and function. Some of these tools include optogenetics, automated patch clamp, and expansion microscopy [1,2]. One goal of this collaboration with Dr. Boyden's Lab is to map the full circuit diagram of mammalian brain, i.e., the connectome.
Today, the only connectome available is for C. elegans. Mapping the connectome is challenging because the typical minimum separation between neuronal membranes, such as at a synapse, is on the order of 20-50 nm. High resolution imaging with an isotropic resolution of at least 50 nm is therefore required. To map the mouse connectome throughout the 0.5 cm3 volume of the brain, we need to acquire a dataset containing 4x1015 voxels, which requires super-resolution methods and data analysis algorithms with unprecedented pixel throughput. Most efforts in connectome mapping are based on electron microscopy. Although electron microscopy has nanometer-scale resolution, image acquisition is much slower than that of optical microscopy. More importantly, automated image processing of greyscale, complex, electron micrographs is difficult, and handling the exabytes of data that would be obtained by electron microscopy of a whole brain is an unmet challenge. We hypothesize that high contrast fluorescent images (where only the neuron branches and synaptic junctions are labeled with high-contrast colors) will overcome the image processing challenges and make the mapping of all the synaptic connections in a mammalian brain possible. This approach demands the development of ultra-high throughput 3D-resolved super-resolution imaging techniques.
LBRC has collaborated with the Boyden lab on several projects over the past two years, making advances in two main areas. First, we have developed a new method [3] for targeted stimulation of a single cell in thick tissue by optogenetic activation of opsins using three-photon temporal focusing (Fig. 1). Second, we have designed and constructed a high-throughput depth-resolved super-resolution RESOLFT system, partly supported by a $200K seed grant from the MIT MINT Program. Instrument construction is currently underway (Fig. 2) that will follow its optimization before biological imaging.
This first-generation super-resolution instrument will be soon ready for testing using cells expressing switchable probes such as Dronpa-M159T or rsEGFP2. The Boyden lab will develop expression /labeling and specimen preparation protocols for brain sections in order to optimize the samples for tracing neuronal dendrites, and identify synaptic clefts accurately, via strategies such as clearing, expansion [2], and inner membrane leaflet anchoring. The instrument design and specimen preparation protocol will be iteratively improved guarantee sufficient stability to undertake the year-long task of imaging a whole mouse brain.
The LBRC will develop a volumetric 3D super-resolution microscope to support Dr. Boyden's connectome mapping project. We will make technology advancements toward increasing imaging resolution, speed, and widening probe selection. If successful, the connectome may provide an underlying ground truth to model information flow in the brain, leading ultimately to an understanding of cognition itself; an important goal of the Brain Initiative / European Brain project. While accurate modeling of whole-brain computation may remain elusive for many years, correlation of functional imaging in vivo with local connectivity information, mapped post-mortem even over a small brain region may inform on the structural origin of many neural pathologies.