170 — Axonal synapse sorting in medial entorhinal cortex

Schmidt et al (10.1038/nature24005)

Read on 06 February 2018
#entorhinal-cortex  #MEC  #synapse-sorting  #webknossos  #connectomics  #EM  #electron-microscopy  #graphs  #mouse 

In my opinion, the biggest challenge in visualizing neural microcircuitry is being able to wrap your head around the enormous density of the data: A single millimeter can hold hundreds of thousands of neurites, and probably hundreds of millions (or more!) of synapses. Ignoring a single voxel can have huge implications when considering the resultant connectome.

One of the most surreal things I’ve ever done is “flown” through dense neuropil using WebKnossos, a tool developed at the Max Planck Institute for Brain Research in Germany. You can watch Flight Mode on YouTube here. In short, it puts you in a (slightly distorted) cavity in the data, and warps the data as you glide through it so that you never feel claustrophobic.

Today’s paper uses a team of 24 tracers — armed with WebKnossos — to understand the patterns of connectivity between neurons in medial entorhinal cortex (MEC). From this interrogation of a few micrometers of mouse cortex, Schmidt et al find a few interesting patterns:

First, excitatory-excitatory connections between neurons in MEC are rare. Instead, inhibitory interneurons likely act as a signal aggregator or filter. This is substantiated by prior electrical recording work.

Second, and more central to the “lesson” of this paper: Synapses self-sort along the length of neurites in MEC. In other words, axons’ inhibitory targets tend to be closer to the soma, whereas excitatory targets tend to be further away. The overall spatial distribution of excitatory or inhibitory synapses is roughly the same; so it is clear that this is not a function of the geometry of the sample, but rather a connectivity pattern common to cells in this brain region.

The authors dub this pattern Path-length-dependent axonal synapse sorting, or PLASS. They hypothesize some possible wiring diagrams for the PLASS circuitry, and demonstrate how PLASS might be a mechanism for tuning spike timing.