Approach
Our experimental approach takes advantage of recent advances in molecular genetic tools and imaging technology that enable targeted labelling, visualization and manipulation of defined populations of neurons in the mammalian brain. Genetically targeted axonal labels reliably reveal patterns of connectivity for single cell types that can be quantitatively assessed throughout postnatal life. Virally-mediated, conditional anterograde and retrograde labeling are now an essential part of our tool kit for visualizing connections between identified populations of neurons. In addition to cellular analysis with confocal microscopy, we have optimized tissue clearing methods and light sheet microscopy, as well as new quantitative 3D data analysis pipelines that depend on implementation of advanced computational methods and machine learning filters. We have also established microendoscopy-based in vivo calcium imaging in freely behaving mice for analysis of brain activity during a variety of behavioral tasks. This collection of visualization and analytical tools enable a direct and quantitative evaluation of circuit architecture, physiology and behavior.
Neuroanatomical analysis of circuit architecture
Created by Dr. Michelle Bedenbaugh
We are currently using conditional (Cre-dependent) AAV-based tracing tools for labeling projections of neurons in either anterograde or retrograde directions. For example, injections of the anterogradely transported AAV (AAV DJ hSyn FLEx mGFP-2A-Synaptophysin-mRuby) into the ARH of a MC3R-cre mouse, transduced neurons to display red/green fluorescence in the cell body, green fluorescence in the axons and red fluorescence in synaptic terminals in targeted regions. This labeling strategy represents a significant advantage for quantitative analysis, because the number of transduced neurons can be counted to normalize the density of labeled terminals, which facilitates accurate comparisons. Similarly, injections of a retrogradely transported AAV (AAVretrog-FLEX-tdTomato) into the PVH conditionally transduces tdTomato fluorescence in neurons that provide direct inputs to a targeted region, with Cre-dependency (only neurons that express CRE). Labeling patterns are analyzed by using confocal microscopy or whole-brain imaging.
Whole-brain imaging
Created by Dr. Michelle Bedenbaugh
In order to take full advantage of these labeling tools and expand our analysis brain-wide, we optimized a whole-brain imaging approach based on tissue clearing and light sheet fluorescence (LSF) microscopy. For this, brains are optically cleared using either iDisco+ or SHIELD protocols and imaged on a selective plane illumination microscope (SPIM; SmartSPIM, LifeCanvas Technologies). The resulting 3D data sets allow evaluation in multiple resampled planes, as well as quantitative comparisons measured in anatomically discrete user defined regions. Similarly, the full distribution of neurons that innervate a target nucleus can be visualized in 3D and the locations of labeled neurons registered to the Allen Brain Atlas Coordinate Reference Framework (CCF, v3.0), by using software developed by LifeCanvas. This imaging pipeline offers enhanced throughput, compared with serial section tomographic methods, while maintaining subcellular resolution and avoiding the need for error-prone section registration algorithms.
Microendoscopy
Created by Dr. Jessica Biddinger, Dr. Dollada Srisai, and Dr. Jose Maldonado
We are actively studying how changes to circuit architecture lead to alterations in postsynaptic signaling in awake behaving mice by using calcium-based microendoscopy. For this, GCaMP6s is genetically targeted to a population of neurons and calcium transients detected through a GRIN lens with a miniaturized microscope. For example, we implemented this deep-brain imaging approach and successfully recorded calcium transients in neurons that express GLP-1 receptors (GLP1R) in the PVH, in the context of food intake and vagal activation (Srisai, et al. Soc Neurosci. Abstr. 2019). We have also integrated microendoscopy in awake behaving mice with quantitative analysis of behavior by using an EthoVision (Noldus) behavioral analysis system to quantitatively measure changes in food intake and accompanying behaviors that are associated with cellular events detected with in vivo calcium imaging.