Constraints - semblancehypothesis.org

How hidden is the solution & how do the constraints help to find it?

The disparate findings from eight domains of the system functions (given after the next figure) are expected to be interconnected through a novel mechanism (Fig.2), analogous to an adaptor protein that binds to multiple molecular partners. This explains how the unique properties of the solution enable large number of features of the system observed from different levels to get integrated in an interconnected manner.

Figure 2. Disparate findings of the system in different levels are interconnected with each other through a unique solution A) The features of the system are detected either directly (represented by the capital letter K) through our sensory systems or indirectly (represented by the capital letters D, E, G, H) through findings such as protein staining or behavioral observations. These features are connected through smaller letters (u, w, v, x), which represent relationships between the features (e.g., the observation of u enables the detection of D). B) By utilizing both commonly employed direct and indirect methods, three clusters of interconnected findings (depicted by dotted lines) are identified across different levels (observations from various fields of brain science). In many instances, it is not possible to establish connections between these clusters. For example, no direct connection was found between: 1) the changes in learning and the internal sensations of memory occurring at millisecond timescales, and 2) sleep & long-term potentiation (LTP). However, by examining the constraints within each cluster, we can assess whether they can be unified through a common operational mechanism. In the context of the nervous system, a vast array of findings and the constraints they provide can be explored. C) By leveraging the constraints from certain features within each cluster of seemingly unrelated findings (e.g. A, B, C), it may become possible to derive a deep underlying principle (a structure-function solution m) that enables their interconnection. This solution is expected to offer a mechanism for generating internal sensations at millisecond timescales. D) The solution m provides an explanation of how various findings within each cluster are interrelated with one another & with findings from other clusters, as shown in B). The ability of solution m to integrate findings across all clusters makes it a further verifiable solution (Fig. from Vadakkan KI (2019) Phys Life Rev.31:44-78.

Constraints (~80) that were used to derive the solution Domain 1 Cellular and Structural
Finding	Constraint (The solution must be able to provide interconnected explanations for all the constraints).
For associating two stimuli & generating motor output reminiscent of the 2^nd stimulus, in response to the 1^st stimulus
All the stimuli during learning & memory retrieval propagate through synaptically-connected neuronal circuits.	Mechanism should operate synchronously with the synaptically connected circuitry.
Any subset of 140 input signals reaching the axonal hillock is sufficient to fire a neuron (Palmer et al., 2014; Eyal et al., 2018).	This extreme degeneracy suggests that the information storage mechanism should be taking place at the level of the input terminals.
Spine neck resistance (Tønnesen et al., 2014) and attenuation of potentials as they propagate along the dendrite towards the soma (Spruston, 2008).	The ideal region for learning changes to take place is the spine head region.
Optogenetic activation of presynaptic inputs in lateral amygdala (LA) forms associative fear memory (Kwon et al., 2014).	LA is a region where inputs from two associated stimuli converge. Hence there should be convergence of inputs at the synaptic level.
Fear learning generates local connectivity between lateral amygdala (LA) neurons (Abatis et al., 2024) as evidenced by depolarization in a small subset of neighboring LA neurons, when a single LA neuron is stimulated.	During cued fear conditioning, lateral connections are expected to occur between LA neurons. These are expected to occur between the dendritic spines of these neurons.
Since associatively learned inputs synapse on to the spines of LA neurons, the inter-neuronal interaction is expected to take place between the synapses.	Since the minimum statement to say that a synapse is activated is generation of postsynaptic potentials, minimum interaction by which this can be accomplished is through inter-spine interaction.
Synapse-dense cortical areas with tightly packed neuropil have adjacent spines (Kasthuri et al., 2015; Zhu et al., 2021; Gemin et al., 2021).	This increases the probability for inter-neuronal inter-spine interactions.
The dendritic arbors of pyramidal neurons display significant territorial overlap (Mizuseki et al., 2011; Bezaire & Soltesz, 2019; Iascone et al., 2020).	This increases the probability of inter-neuronal inter-spine interactions.
The sister branches on a neuron’s dendritic tree often avoid overlapping (Grueber & Sagasti, 2010).	This increases the probability of inter-neuronal inter-spine interactions.
The mean inter-spine distance on the dendrite of a pyramidal neuron exceeds the mean spine diameter (Konur et al., 2003).	This allows interaction between the spines that belong to different neurons.
Over half of the spine surface area lacks ensheathment by astrocytic processes (Ventura & Harris, 1999).	The remaining half of the spine surface area is free to interact with the abutted spine surfaces of other neurons.
Synapses devoid of astrocytic coverage emerge in the amygdala during the consolidation of Pavlovian threat conditioning (Ostroff et al., 2010).	The reduction or disappearance of astrocytic pedicels during the consolidation may increase the probability of inter-neuronal inter-spine interaction.
Most excitatory glutamatergic synapses are located on dendritic spines, which enlarge during learning. Glutamate induces spine enlargement in both hippocampal slices (95%) (Matsuzaki et al., 2004) & the neocortex in vivo (22%) (Noguchi et al.,2019).	Spine enlargement can both increase the surface area and also displace the extracellular matrix (ECM) increasing the probability for inter-neuronal inter-spine interaction.
Fear conditioning is associated with enlarged synapses on the dendritic spines of LA neurons (Ostroff et al., 2010; Choi et al., 2021).	Synapse enlargement can result from the expansion of either pre- &/or postsynaptic terminals. Since inter-spine interaction is an anticipated learning change, this matches.
Motivation enhances learning & is associated with the release of dopamine, which activates dopamine receptors in various regions of the brain (lino et al., 2020).	Dopamine is known cause spine expansion (Yagishita et al., 2014). Abutted spines from different neurons, when expanded can undergo certain inter-spine interactions.
Contextual fear conditioning recruits newly synthesized GluA1-containing AMPA receptors into the spines of hippocampal memory-ensemble cells in a learning-specific manner (Matsuo et al., 2008).	Exocytosis of vesicles adds vesicle membranes segments to the spine membrane increasing spine size, which can facilitate inter-spine interactions.
Autophagy leads to memory destabilization & erasure of auditory fear memories, a process associated with AMPA receptor endocytosis (Shehata et al., 2018).	Endocytosis of vesicles removes vesicle membranes segments from the spine membrane decreasing spine size, which can separate interacting spines of different neurons.
An inhibitor of AMPA receptor endocytosis partially rescued long-term memory deficits in mice with elevated levels of amyloid-β (Yan et al., 2024).	Preventing endocytosis of vesicles prevents removal of vesicle membranes segments from the spine membrane & prevent decrease in spine size, maintaining abilities for inter-spine interactions.
Mice injected with histone acetyl transferase (HAT) exhibit enhanced fear memory. Neurons in which HAT was overexpressed are part of the engram (Santoni et al., 2024). Removal of histone protein from DNA enhanced fear memory. The study found that a) neurons in which HAT is over-expressed are the neurons that fire during memory retrieval, & b) optogenetic silencing of these specific set of neurons prevents fear memory recall.	Removal of histone proteins from DNA facilitates the gene expression, whose protein products can be used for the inter-neuronal inter-spine interaction mechanism. HAT over-expressed neurons allow more inter-spine interactions to form during learning. These excess inter-spine interactions help to generate robust first-person property of fear and behavioral motor actions reminiscent of fear.
Most learning events result in a working memory that lasts only for a short period. All long-term memories has had working memories immediately after learning.	Inter-spine interaction is a highly reversible process. Inter-spine interactions extending to inter-membrane hemifusion events can stabilize the inter-spine interactions variable periods.
The capacity to store extensive sets of learning-induced changes underlies the ability to retrieve a large number of behavioral expressions during memory retrieval events. Using a finite number of dendritic spines this must be achieved.	A single spine that can interlink with more than one spine can lead to the formation of islets of interlinked spines (IILSPs). This can allow a specific cue stimulus can reactivate specific sets of interlinked spines to generate corresponding sets of behaviors.
Higher brain functions occur only when the frequency of oscillating extracellular potentials falls within a narrow range, as evidenced by EEG recordings (Rusalova, 2006).	Oscillations needs vector components that contribute to it. Since all cortical neurons have their apical dendritic terminal attached to the inner pial surface, large number of inter-spine interactions can be formed. Synaptic transmission in mostly vertically oriented synapses & horizontally oriented inter-spine interactions can contribute vector components for oscillations.
Even though extracellular recordings like EEG is biased by filtering effects (e.g. extracellular tissue act as a low-pass filter), brain operates in a narrow range of frequencies of EEG waveforms recorded from specific locations. For e.g. alpha rhythm (8–12 Hz) recorded from occipital cortex during relaxed wakefulness (Niedermeyer & da Silva, 2017).	The operational mechanism is expected to generate vector components of the oscillating extracellular potentials, accompanied by corresponding intracellular ionic concentration changes within the neuronal processes of the involved neurons. Inter-spine interaction mechanism provides vector components for oscillating potentials.
A disconnect between dendritic depolarization & neuronal firing has been observed during fear conditioning (d’Aquin et al., 2022).	An operational mechanism, most likely related to the first-person property, is expected to emerge at the dendritic level, independent of neuronal firing.