Just had the thought that we’ve traditionally thought about neurotransmitters as the signalling mechanism between neurons as part of the neuron centric view of brain function.
How would brains look if we viewed these chemicals as control/relay agents which enabled local signalling between distinct glial populations? Does each transmitter allow system level signalling between disparate glial populations?
What would this tell us about brain function? Perhaps GABA/GLU refers to general dorsal/ventral signalling regime? This has some pretty interesting consistency with the assumed function of our valence centers (amygdala complex/habenula complex, etc) which all seem to work with this GABA/GLU balance to generate effect.
Just as importantly, it draws us back to the hypothalamus/mammillary bodies as fairly critical modifiers of behavior.
It seems possible that how neurons are connected (e.g. connectome maps) might have far less of an impact on function than overall communication efficiency between populations of glia on each distinct signal channel. Someone may have highly efficient GABA signalling, but low dopamine signalling efficiency. This would result in a specific behavioral predisposition. The efficiency would be a product of maybe astrocyte clearing/generation rate, receptor/transmitter density both glia and neurons, neuronal transmission efficiency (myelination, electrical properties), and population density.
It’s interesting that under this construct we get a more nuanced explanation for heavy synaptic pruning during post natal brain development, as preferred circuit paths are established local populations will reduce redundant connections to reduce noise (and energy use) between local groups. At least in the cerebral side, we “learn” by creating maps of observed data and pruning away the redundant. As astrocytes are responsible not only for this pruning, but the modification of spines themselves to store the data, it seems like children essentially have an open map and as we age we prune away the possible.
Going back to our “autism” map, we definitely have phenotypes which do not prune as well as phenotypes which over prune. What does this look like as far as presentation? Will need to refresh to make sure I’m remembering correctly.
I’m getting fairly certain that we can tie g to astrocytic metabolic efficiency, using metabolic efficiency as an analog for signalling efficiency. We should see an inverse relationship between energy use and g if this assumption is valid.
I have like a billion concurrent threads going on right now so I need to stop and reconcile some of them (and try not to space out too much), but abandoning the neuron centric model does provide some extremely intriguing new paths for both research and idle pondering.
What are some possibilities I’m missing?