Imaging the Connectome
Jeff W. Lichtman, MD, PhD
Friday, October 26, 2012
Tweet
Transcript
Embed
Email
Bookmark
<<
<
>
>>
Tweet
Facebook
Email
Chapters
Introduction
Structure and Function of Nervous System is Complicated
Brain Organized over Sizes that Span 6 Orders of Magnitude
Engramitization
Connectomics
The Golgi Technique
Cajal’s Drawing of a Circuit in the Cerebral Cortex
The Sparse World View: Canonical Circuits
The Saturated Connectome World View
Strategy to Label with Many Different Colors for Saturated Connectomic Reconstructions
Use Color to Disambiguate Neurons Using Strategy 1: 'Brainbow'
Astrobow (Mostly 3 Color)
Color Combinations via Tandem Copies to Generate >3 Colors
Hippocampus
Dentate Gyrus
GenII Brainbow in Cerebellar Purkinje Cell Dendrite
Three Images Suffices for Many Colors
Gen II Membrane-expressed Brainbow Mice
One Axon per Neuromuscular Junction in Adult
Interscutularis Connectome
Wiring Sub-Optimality in the Mammalian Nervous System
Competition for Synaptic Sites
All to All v. Sparse
Illustrating Several Seasons in an Axon’s Life
At Birth, Junctions Are Weakly Innervated by Individual Axons
All Inputs at Each Neuromuscular Junction at Birth
10 or More Axons at Each Neuromuscular Junction at Birth
Neonatal Connectomics
No Topography: Every Axon Goes Nearly Everywhere
Omohyoid 1: 411 Neuromuscular Junctions
DevelopingConnectome:Non-Random Graded Connectivity
Linear Plan
A Circuit Motif: Linear Order Connectional Matrix
The Neonatal Wiring Diagram Structure per se Reveals the Experience That Set it Up
How about the Cerebral Cortex?
All Axons and Dendrites: Too Much Stuff
BrainStorm?
Multi-Color STOM Imaging of Neurons
AAV-Membow for Subset Labeling
Inhibitory Axons Only
But What if We Want to See All the Connections?
Automatic Tape-collection Mechanism for Ultramicrotomes
Throughput: ~1000 Sections/Day; 30nm
Strategy to Make Library
1 Wafer180 Sections
55 Wafers = 10,000 Sections
WaferMapper Software
Tape on Silicon Wafers in Scanning Electron Microscope
Thin Section Scanning Electron Microscopy
30 nm/Pixel
Computer-assisted “VAST” Saturated Hand Segmentation of Cortex
Overlap of the Dendritic Spine Fields of Nearby Dendrites
We Began 3+ Years Ago...
State of the Art: Lateral Geniculate of Thalamus
Stitched 100k x 100k Mosaic at 4nm per Pixel
Next Few Months….
40M Pixels/s (Secondary Electron Imaging)
Next 1-2 Years…
A 61 bean PrototypeScanning Electron Microscope
Second Problem: Turning Images Into Data
But What about the Segmentation Problem?
Fully Automated Segmentation by Bipartate Matching with Fusion
Automated v. Manual
'Saturated' Reconstruction of All Input
The 113 Axons that Innervate All the Spines in a Small Segment of the R, G and B Apical Dendrites
Synaptic Clustering: A Motif?
In Situ Synaptic Architecture
Acknowledgements
No transcript for this webinar
More Information
Request info