The prevailing understanding of how synapses work is driven by a fairly low resolution understanding of cellular mechanics.
Rather than “neuron to neuron” communication, or even “tripartite” communication, imagine the synapse as having thousands of individual peptide communication channels, each which transmit or internalize a different signal.
Traditional neurotransmitters do not contain messages themselves, instead they allow a much more generalized activation signal. The neurotransmitter signal adjusts the sensitivity of a cell to those incoming peptide messages.
Dopamine for instance adjusts sensitivity to certain classes of peptides (like neurotensin), while a serotonin signal may adjust sensitivity to interleukins.
The actual effect is that these transmitters modify the methylation rates of RNA generated in response to peptide receptor binding.
Tree complexity is an artifact of complexity of peptide receptor transmission/binding points, rather than the morphology of the cell needing to modify an electrical signal to certain properties.
A typical inter synaptic pulse will consist of hundreds of different peptides to complete a successful message.
The transmission/reception points are modulated by astrocytes, which monitor the messages and remodel the spines to generate binding points necessary to store the information relevant to it’s local context. Tree morphology should be correlateable to information processing density of a cell.
Examining this in the context of memory as a whole, the actual data in a local group is preserved in the specific peptide receptor combinations expressed in parts of the synapse. That unique combination of peptide expression is literally a micro engram, a bit of a memory, which when integrated with input from several neurons forms a discrete memory chunk.
An example that sprang to mind earlier is that these peptides are the words of intercellular communication and the specific combination/interaction of these peptides form the language, which are the basis of behavior/memory. Traditional neurotransmitters work like intonation or context specific non-linguistic modifiers to language (saying a word quietly or loudly sometimes imparts different context for instance).
This integrates a few important concepts. This illustrates why astrocytes are necessary for dynamic information acquisition, as they contain the ability to generate distinct RNA phrases, and these phrases are expressed into peptide receptors on neurons.
This implies that the receptor environment in brains is far more complex than we understand right now, and are probably just as unique as genetic expression. Which is what we see in practice.
It’s important to remember that even though we think about cells in terms of bodies and networks, each individual cell is a discrete environment. Much like individual humans attempting to perceive the external social environment, they have no direct access to the internal environment of other cells. In order to coordinate any type of behavior, cells must be able to interpret and conform themselves to external information.
The only information cells have about the external environment comes from one the sensory channels – chemical sensing, mechanical sensing, or “energy/EM” sensing. Heart cells for instance can use both chemical and mechanical sensing to stay in rhythm with other heart cells, they use chemical sensing to understand which external mechanical rhythms they should be “paying attention” to.
A significant feature of the gap is that it allows synchronization of signals across many different external sources, creating a synchronization mesh instead of a chain. This mesh is the nature of “wave” signaling we see in brains.