PeriNeuronal Nets: Protein-Sugar Structures Surrounding Neurons

[BillTre]

This article from the Scientist (which I think is open access), describes periNeuronal Nets. These are protein and sugar molecular assemblies which surround neurons (maybe only some).
They seem to be involved in the neuron cell biology underlying synaptic plasticity (which in turn underlies learning, making memories, preserving memories, losing memories, etc., things involved with changes in synapse onto these neurons).
They molecular composition is described along with possible involvement in diseases/treatments.

 

[Pythagorean]

Perineuoronal nets have been implicated in "critical windows", whereby there are developmental phases in life where learning certain things is optimal due to the lack of perineuronal nets in a specialized brain region. Once a certain phase of development is reached, perineuronal nets form around a synapses in a region, preventing further synaptic change to the ensemble.

 

[Gary Feierbach]

That would imply that learning in that region is connection based and does not involve varying synaptic weights. A couple of the latest papers also indicate that synaptic weights are irrelevant (they are either connected or not connected) and the real neuronal business (a single weight changing) is done in each dendritic node joining the neural body.
See: "Adaptive nodes enrich nonlinear cooperative adaption by links" by Shira Sardi, et al.

 

[Pythagorean]

That would imply that learning in that region is connection based and does not involve varying synaptic weights. A couple of the latest papers also indicate that synaptic weights are irrelevant (they are either connected or not connected) and the real neuronal business (a single weight changing) is done in each dendritic node joining the neural body.
See: "Adaptive nodes enrich nonlinear cooperative adaption by links" by Shira Sardi, et al.

On the surface, that seems consistent with studies using "electrotonic length". Dendrites can vary their effective length (in terms of amplitude attenuation) somewhat independently of their actual length based on morphology (or possibly both structure and function).

In this example, they're looking at how electrotonic length changes with age. It's the only paper I could find that wasn't behind a paywall, but I seem to remember learning it as a general property of neurons and plasticity:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2742588/figure/fig9/

But I wouldn't be so quick to throw out synaptic weights all together. Homeostasis with multiple pathways is always tricky business.