Findings after the hypothesis & their new explanations

Current studies use behavior (in lieu of first-person properties of brain functions) to examine higher brain functions. To verify the semblance hypothesis, it is necessary to examine experimental findings obtained by using behavior. Experimental results are bedrocks of science. The aim of searching for retrodictive evidence is to reexamine whether alternate explanations become possible for the findings in these experiments. It is necessary that these alternate explanations are interconnected and match with the solution derived by the semblance hypothesis using constraints from previous observations from the system.

A. In physiological conditions & artificial systems

Chromatin plasticity predetermines neuronal eligibility for memory trace formation. Santoni G, Astori S, Leleu M, Glauser L, Zamora SA, Schioppa M, Tarulli I, Sandi C, Gräff J. Science. 2024 Jul 26. Article

Santoni et al. found that neurons in which histone acetyl transferase (HAT) was overexpressed (& that led to increased histone acetylation) are neurons in the engram. Histones are proteins covering DNA. HAT is non-specific regarding the locations from where it removes histones over the DNA sequences that constitute the genes. Histone acetylation by HAT decreases the affinity of histones to DNA, exposing all the genes, which will facilitate transcription. However, in a state of homeostasis where certain specific genes need to be expressed, HAT overexpression is expected remove histones over those genes. Hence, in the case of fear conditioned learning, HAT is expected to promote expression of genes that are essential for learning changes necessary for memory storage. 

Using the findings in this paper, it is possible to ask, “Is it providing any interconnected explanation for the semblance hypothesis?” Since IPLs are expected to be formed during learning between abutted spines (of different neurons or different branches of a single neuron), to retain the IPLs for maintaining memory of a survival requiring action, the neuronal cell will restrict using membrane segments from those spines for endocytosis. This can be expected to result as an evolutionary adaptation. Hence, more phospholipid molecules are expected to be synthesized to make lipid membrane segments to retain the formed IPLs. This leads to homeostatic mechanisms to synthesize phospholipids by fatty acid synthesis (mainly palmitic acid) by a multi-enzyme complex consists of seven different enzymes followed by desaturase and elongase enzyme actions, synthesis of phospholipids, their transport and incorporation into the plasma membranes. In these contexts, HAT is expected to remove histones from the corresponding DNA sequences to facilitate expression of a specific set of enzymes. 

In summary, HAT injected mice exhibited stronger fear memory because those IPLs that were formed during fear learning, were maintained. It is reasonable to expect that this became possible since there was no homeostatic pressure for their reversal since enzymes along the path of making phospholipid membranes were expressed in sufficient quantity by  HAT.

Authors found increased expression of synaptic proteins. Since IPLs are formed between spines that are part of synapses, naturally it will lead to several homeostatic mechanisms to replenish substrates to maintain existing synapses whose spines formed specific IPLs

Optogenetic silencing of the epigenetically altered neurons prevented fear memory recall. Since we are examining behavior, silencing these neurons will prevent their firing, which will not allow firing of downstream neurons to manifest behavioral motor actions (foot withdrawal) during fear memory retrieval.

Note: It is highly necessary to start experiments by making different presuppositions. For example, the present work viewed that memories are encoded by sparse populations of neurons. To understand how memories are stored, it is necessary to view memories as a first-person property and derive a testable mechanism (e.g. the semblance hypothesis) using constraints from a large number of findings in different experiments and test whether derived solution still hold true.

Higher-order interactions between hippocampal CA1 neurons are disrupted in amnestic mice. Yan C, Mercaldo V, Jacob AD, Kramer E, Mocle A, Ramsaran AI, Tran L, Rashid AJ, Park S, Insel N, Redish AD, Frankland PW, Josselyn SA. Nat Neurosci. 2024 Jul 19. Article.

Authors found that TAT-GluA23Y, an inhibitor of AMPA receptor (AMPAR) endocytosis, partially rescued long-term memory deficits in mice with high levels of amyloid-β. The semblance hypothesis has put forward formation of (inter-postsynaptic (inter-spine) functional LINKs (IPLs) as the engram mechanism during learning and reactivation of these IPLs as the engram mechanism during memory retrieval generating first-person inner sensations. Endocytosis of AMPARs utilizes membranes segments from the lateral spine membranes to form endosomes. This will lead to a reduction in the size of the spine heads and lead to reversal of the IPLs formed during learning (Fig.1). By inhibiting AMPAR endocytosis, TAT-GluA23Y will maintain the spines in their maximum possible sizes that will facilitate IPL formation. Hence, even though amyloid-β within the extracellular space hinder IPL formation, TAT-GluA23Y can facilitate formation of more IPLs. This will enable partially rescue of long-term memory deficits in mice with high amyloid-β.

 
Figure 1. Figure showing how endocytosis will cause reduction in the size of the dendritic spine and reverse newly formed IPL. A) Cross section through two dendritic spines that are inter-LINKed to form a hemifused structure. B) Membrane segments invaginate from the spine membranes to form endosomes. In this process, the circumference of the spines reduces pulling the IPLs to separate. Here the hemifused membranes reverse back to the stage of abutted membranes. C) When the endosomes are formed by using membrane segments from the spine membranes, IPLs completely reverse back to form independent spines. Note that endosome membranes are made of part of the membrane region that was forming the IPL in Fig.A. Red: inner membrane segments of the spines become outer membrane segments of the endosomes. Blue: outer membrane segments of the spines become inner membrane segments of the endosomes.

Authors also found that learning causes an increase in dendritic spine density and dendritic spine head diameter even in mice with high levels of amyloid-β. Based on the semblance hypothesis, increase in spine density can increase the probability for spines that belong to different neurons to form IPLs (this facilitates associative learning between stimuli that generates different motor outputs) or to form IPLs between different branches of the same neuron (this facilitates associative learning between stimuli that generate same motor output).

Finally, by comparing the accuracy of different classifiers trained on the neuronal Ca²⁺ data to decode freezing behavior, authors concluded that certain complex interactions between hippocampal CA1 neurons are necessary for normal memory. Based on the semblance hypothesis, these complex interactions between CA1 neurons of the same neuronal order occur due to the formation (during learning) and reactivation (during memory retrieval) of IPLs. Interconnections of several spines lead to the formation of islets of inter-LINKed spines (IILSs). The engram consists of relationships between the spines within the IILSs.

Barcoding of episodic memories in the hippocampus of a food-campus of a food-caching bird. Chettih SN, Mackevicius EL, Hale S, Aronov D (2024). Cell. 187(8):1922–1935.e20. Article

Hippocampal neurons in chickadees show patterns of activity of CA1 neurons that are specific to the locations of hidden food resulting from retrieval of specific memories but are independent of place fields (set of CA1 neurons that fire when the bird reaches a place) for the same locations. In other words, one set of CA1 neurons fire during memory retrieval associated with reaching a place and another set of CA1 neurons fire when the bird reaches a particular location. Authors call these different sets of CA1 neurons that fire as barcodes.

It was previously shown that place cells (hippocampal CA1 neurons) (O’Keefe and Dostrovsky, 1971) are also associated with different higher brain functions such as trajectory of motion (Frank et a., 2000); location and memory in different contexts (Pastalkova et al., 2008), reward (Gauthier and Tank, 2018), auditory frequency in cognitive maps (Aronov et al., 2017), and abstract learned variables (Schuck and Niv, 2019; Park et al., 2020). In humans, visual images led to firing of sparsely located hippocampal neurons among a larger population of non-responsive neurons (Waydo et al., 2006). Firing of different sets of sparse hippocampal pyramidal neurons during different brain functions necessitates a mechanistic explanation. There is a combinatorial mechanism operating at the input levels of these CA1 neurons. To arrive at this mechanism, it is possible to view the following.

1. The different brain functions associated with separate operations operate at a lower level than the CA1 neurons – this can be at the synaptic levels between CA3 & CA1 &/or CA1 & CA3, &/or CA3 & granule neurons &/or CA3 & entorhinal neurons.

2. Since any set of nearly 140 input signals can fire a cortical pyramidal neuron, pyramidal neurons having thousands of input terminals can be fired by a gigantic combination of input signals arriving through those input terminals (Palmer et al. 2014; Eyal et al., 2018).

3. Reaching a place is associated with first-person inner sensations of that place. Memory of food located in a specific place is associated with first-person inner sensation of food in a particular location.

4. Firing of sets of sparsely located CA1 neurons indicates that specificity to any function is achieved by firing of certain CA1 neurons being held at subthreshold activation states.

These lead to the question, “What type of a mechanism occurring at the lower level, presumably at the input level of CA1 neurons that can generate both first-person inner sensations and firing of different sets of CA1 neurons?” Semblance hypothesis provides a solution to this. Semblance hypothesis describes an operational mechanism at the initial locations of the input terminals (see the FAQ section on the website). To summarize, according to the semblance hypothesis, learning generates inter-neuronal inter-spine LINKs between spines that belong to different neurons at the locations of convergence of pathways through which associatively learned stimuli propagate. Two inter-LINKed spines can inter-LINK with more spines of different neurons during additional associative learning events to form an islet of inter-LINKed spines (Fig.1).

Figure 1. One islet of inter-LINKed spines having 32 spines (dendritic spines) that belong to 32 different neurons are shown. Due to space limitations only neurons (N1, N2, N3 and N4) of 4 inter-LINKed spines are shown in the picture. Depending on the location of the inter-LINKed spine where the input signals from a cue stimulus arrive in an islet of inter-LINKed spine, its neighboring inter-LINKed spines will be depolarized. This also depends on whether the neighboring spines are receiving inhibitory inputs at the same time. Along with this, depolarization propagates from the inter-LINKed spines to their neurons. If the neuron is being held at sub-threshold activate level short of the potentials that are arriving, then the neuron will fire.

Note that spines of one neuron can be part of different islets of inter-LINKed spines (Fig.2). Chettih et al.,’s work (Chettih et al., 2024) shows that at the time of memory retrieval of a place or the location, different sets of neurons fire. According to semblance hypothesis, during different learning events, converging pathways of associated input signals from inter-LINKed spines. Different spines of a neuron can inter-LINK with spines that belong to different neurons. Since an inter-LINKed spine can inter-LINK with more than one spine, it leads to the formation of islets of inter-LINKed spines. When this continues, spines of a neuron will be part of different islets of inter-LINKed spines, which will give a picture shown in Fig.2.

                       
Figure 2. Spines of a neuron are part of different islets of inter-LINKed spines. Here, three islets of inter-LINKed spines are shown. Note that one spine of the neuron N1 on the leftmost side is part of the leftmost islet of inter-LINKed spines (having 19 inter- LINKed spines. Also, another spine of the same neuron N1 is part of the middle islet of inter-LINKed spines (having 31 inter-LINKed spines). Similarly, neuron N4 has one spine as part of the middle islet of inter-LINKed spines. It also has another spine that is part of another inter-LINKed spines (with only 8 inter-LINKed spines). When a cue stimulus arrives at an inter-LINKed spine in an islet of inter-LINKed spines, then units of first-person inner sensations is expected to be generated. Along with this, potentials from the inter-LINKed spines reach their neurons. If addition of these potentials can allow the net potential to cross the threshold for firing an action potential, then that neuron fires.

According to the operational mechanism derived by the semblance hypothesis, first-person inner sensation of a memory of food or a place occurs at the inter-LINKed spines within the islets of inter-LINKed spines. Furthermore, firing of postsynaptic neurons depends on amount of depolarization arriving at these sub-threshold activated neurons. Based on these, findings in Chettih et al.,’s work (Chettih et al., 2024) can be explained in terms of specific sets of input signals arriving at different sets of islets of inter-LINKed spines leading to firing of different sets of neurons that are being held at sub-threshold activation levels. Each higher brain function in which hippocampal CA1 neurons are involved, it leads to firing of different sets of neurons that in turn lead to the appearance of a barcoding of firing neurons.

Chettih SN, Mackevicius EL, Hale S, Aronov D (2024) Barcoding of episodic memories in the hippocampus of a food-campus of a food-caching bird. Cell. 187(8):1922–1935.e20. Article

Eyal G, Verhoog MB, Testa-Silva G, Deitcher Y, Benavides-Piccione R, DeFelipe J, de Kock CPJ, Mansvelder HD, Segev I (2018) Human cortical pyramidal neurons: From spines to spikes via models. Front. Cell Neurosci.12:181. PubMed

Frank LM, Brown EN, Wilson M (2000) Trajectory encoding in the hippocampus and entorhinal cortex. Neuron 27:169–178. Article

Gauthier JL, Tank DW (2018 ) A dedicated population for reward coding in the hippocampus. Neuron. 99(1):179–193.e7. Article

O’Keefe J, Dostrovsky J (1971) The hippocampus as a spatial map. Preliminary evidence from unit activity

 in the freely-moving rat. Brain Res. 34:171–175. Article

Palmer LM, Shai AS, Reeve JE, Anderson HL, Paulsen O, Larkum ME (2014) NMDA spikes enhance action potential generation during sensory input. Nat Neurosci 17(3):383-390. PubMed

Park SA, Miller DS, Nili H, Ranganath C, Boorman ED (2020) Map Making: Constructing, Combining, and Inferring on Abstract Cognitive Maps. Neuron 107:1226–1238.e8. Article

Pastalkova E, Itskov V, Amarasingham A, Buzsáki G (2008) Internally generated cell assembly sequences in the rat hippocampus. Science. 321(5894):1322–1327. Article

Schuck NW, Niv Y (2019) Sequential replay of nonspatial task states in the human hippocampus. Science 364:eaaw5181. Article

Waydo S, Kraskov A, Quian Quiroga R, Fried I, Koch C (2006) Sparse representation in the human medial temporal lobe. The Journal of neuroscience. 26:10232–10234. Article

A ubiquitous spectrolaminar motif of local field potential power across the primate cortex. Mendoza-Halliday D, Major AJ, Lee N, Lichtenfeld MJ, Carlson B, Mitchell B, Meng PD, Xiong YS, Westerberg JA, Jia X, Johnston KD, Selvanayagam J, Everling S, Maier A, Desimone R, Miller EK, Bastos AM (2024) Nat. Neurosci. doi: 10.1038/s41593-023-01554-7. Article

During the development of cerebral cortex, there is a particular pattern in which neuronal cells move and settle in the cortical layers. This has a profound influence on the inter-neuronal inter-spine interactions and hence the waveform of oscillating extracellular potentials. At one stage of development, neuronal cells from the periventricular zone move up along the vertically oriented processes of the radial glia towards the pial surface (Marín-Padilla, 1998). The neurons that reach the subpial region anchor their processes to the marginal zone close to the pia and then descend towards the direction of the ventricular zone area (Fig.1). As the new neurons arrive at their destination, they continue to settle one above the other. Thus, the first set of neurons becomes the sixth neuronal layer of the cortex. This is followed by the fifth neuronal layer and so on. Layer 1 cortical neurons that are mostly GABAergic send horizontal processes interconnecting several postsynaptic terminals of apical tufts. Because of the subpial anchoring of the apical dendritic tufts for all the neurons of all the layers, dendritic arbors of all cortical pyramidal neurons overlap each other more densely in cortical layer 2. The overlap will be less in layer 3; further reduced in layer 4 and so on (Fig.2). Maximum number of synapses and hence inter-neuronal inter-spine interactions are expected to occur in layer 2 providing large number of components of oscillating extracellular potentials in layer 2. Furthermore, large number of thalamocortical inputs also contribute large number of components for these oscillations. Hence, the net effect is expected to increase the frequency of oscillating extracellular potentials in layer 2 when two differential electrodes (extracellular) are placed at two locations in the cortex. This provides an explanation for the observation of high frequency gamma waveforms of oscillating extracellular potentials in layers 2 & 3, and comparatively low frequency alpha and beta waves in deeper layers in the work of Mendoza-Halliday et al.,. This also acts as an indirect retrodictive finding that matches with the semblance hypothesis.

Figure 1. Stages of neuronal migration and arrangement of neurons in different neocortical neuronal orders. Both figure panels are views of the vertical section through the cortex. A: Progenitor neuroepithelial cells in the fetal ventricular zone proliferate, and the newly formed daughter cells migrate along the processes of the radial glia towards the superficial layer of the cortical plate. The new cells develop processes, and the dendrites are anchored to the extracellular matrix structural proteins at this region. As new cells arrive at the superficial layer, the older ones get pushed towards the direction of the ventricular zone. However, their apical dendrites remain anchored to the superficial layer. Since the dendrites are already anchored to the superficial layer, the main dendritic stem elongates. This continuous process results in the displacement of the oldest cell layer, namely layer 6 located close to the ventricular zone and the last arrived cells to remain at the most superficial layer as neuronal order 1. B: Figure showing the dendritic trees of neurons that belong to different neuronal orders (numbered 1 to 6). Note that the dendrites of neurons of almost all the neuronal orders anchor at the subpial region, making this region rich in dendritic spines. As the dendritic spine density is very high at the cortical layers 1 and 2, the role of this arrangement contributing to the interspine interactions and oscillatory waveform of the cortical surface-recorded potentials are explained in the following sections (figure modified from Vadakkan, 2015).

Figure 2. Anatomy of pyramidal neurons and locations of spike generations. A: Diagram showing different locations of neuronal processes that are capable of producing regenerative potentials (spikes). These include apical tuft, apical trunk, oblique, basal, and axon. Synapses in distal dendrites produce EPSP of amplitude more than 10 mV; whereas those proximal to the soma produce EPSP amplitude of 0.2–0.3 mV. The EPSPs from distal dendrites attenuate from nearly 10 millivolts (mV) to nearly 0.014 mV (more than 900-fold attenuation) as they reach the soma (Spruston, 2008). B: Diagram showing the locations of spike generation and inhibitory mechanisms to regulate spike propagation. In the apical tuft, oblique, and basal dendrites, several dendritic conductance contributed by regenerative NMDA receptor current trigger dendritic plateau potentials with a rapid initial sodium spikelet followed by a plateau phase that collapses abruptly (Schiller et al., 2000). At the apical trunk, calcium spikes are generated. Axonal spikes (action potentials) are generated at the axon initial segment. Inhibitory inputs can regulate the net potential reaching the soma. Both recurrent collateral and thalamo–cortical inputs control the generation and propagation of the spikes at different locations. Layer 1 cortical neurons that are mostly GABAergic send horizontal processes interconnecting several postsynaptic terminals of apical tufts. Representative traces of different spikes are shown (figure modified from Vadakkan, 2015).

Mendoza-Halliday D, Major AJ, Lee N, Lichtenfeld MJ, Carlson B, Mitchell B, Meng PD, Xiong YS, Westerberg JA, Jia X, Johnston KD, Selvanayagam J, Everling S, Maier A, Desimone R, Miller EK, Bastos AM (2024) A ubiquitous spectrolaminar motif of local field potential power across the primate cortex. Nat. Neurosci. doi: 10.1038/s41593-023-01554-7. PubMed

Spruston N (2008) Pyramidal neurons: dendritic structure and synaptic integration. Nat. Rev. Neurosci. 9(3):206–221. PubMed

Vadakkan KI (2015) A pressure-reversible cellular mechanism of general anesthetics capable of altering a possible mechanism for consciousness. SpringerPlus 4:485. doi: 10.1186/s40064-015-1283-1. PubMed

Schiller J, Major G, Koester HJ, Schiller Y (2000) NMDA spikes in basal dendrites of cortical pyramidal neurons. Nature 2000;404:285–289. PubMed

Marín-Padilla M (1998) Cajal–Retzius cells and the development of the neocortex. Trends Neurosci. 21(2):64–71. PubMed

Vadakkan KI (2016) Rapid chain generation of interpostsynaptic functional LINKs can trigger seizure generation: Evidence for potential interconnections from pathology to behavior. Epilepsy Behav. 59:28–41. PubMed

Hotspots of dendritic spine turnover facilitate clustered spine addition and learning and memory. Frank AC, Huang S, Zhou M, Gdalyahu A, Kastellakis G, Silva TK, Lu E, Wen X, Poirazi P, Trachtenberg JT, Silva AJ. Nat Commun. 2018 Jan 29;9(1):422. PubMed

See supplementary figure 8 in the above article. Learning leads to loss of spines & formation of new spines at those regions (spine turnover). Why should spines get lost? Based on the semblance hypothesis, learning leads to inter-neuronal inter-spine interaction leading to inter-postsynaptic functional LINKs (IPLs) (see figure 8 in FAQ section of this website). Inter-spine fusion is at the extreme end of this spectrum of changes. The nature of IPLs depends on several factors. These include a) nature of fatty acids in the phospholipid molecules that form spine membranes, and b) intensity of stimuli that affect propagation of signals towards the IPLs. If IPL formation leads to an extreme change of inter-neuronal inter-spine fusion, then it will lead to mixing of the contents of cytoplasm of two neurons. Since even adjacent neurons of a similar type vary in their gene expression/protein content (Kamme et al., 2003; Cembrowski et al., 2016), it is reasonable to expect cellular mechanisms for closing the fusion pore. If it is not possible, then the neurons will trigger mechanisms to remove the spines. This can explain spine loss. As a homeostatic mechanism, the involved neurons can be expected to produce new spines using phospholipids that resist inter-spine fusion. Thus, the basic operational mechanism of semblance hypothesis can be extended to provide a mechanistic explanation for spine turnover during learning.

Cembrowski MS, Bachman JL, Wang L, Sugino K, Shields BC, Spruston N (2016) Spatial Gene-Expression Gradients Underlie Prominent Heterogeneity of CA1 Pyramidal Neurons. Neuron. 89(2):351-368. PubMed]]

Frank AC, Huang S, Zhou M, Gdalyahu A, Kastellakis G, Silva TK, Lu E, Wen X, Poirazi P, Trachtenberg JT, Silva AJ (2018) Hotspots of dendritic spine turnover facilitate clustered spine addition and learning and memory. Nat Commun. 29;9(1):422. PubMed

Kamme F, Salunga R, Yu J, Tran DT, Zhu J, Luo L, Bittner A, Guo HQ, Miller N, Wan J, Erlander M (2003) Single-cell microarray analysis in hippocampus CA1: demonstration and validation of cellular heterogeneity. J Neurosci. 23(9):3607-3615. PubMed

Synapse-specific representation of the identity of overlapping memory engrams. Abdou K, Shehata M, Choko K, Nishizono H, Matsuo M, Muramatsu SI, Inokuchi K (2018) Science. 360(6394):1227-1231. (Unfortunately, I saw this paper only on the 20^th March 2023. So, posting this explanation very late). Article

Related article: Entorhinal cortex directs learning-related changes in CA1 representations (Grienberger and Magee, 2022) Nature. November, doi: 10.1038/s41586-022-05378-6. Article

Semblance hypothesis has provided a mechanistic explanation for both memory (inner sensation of features of the item/event whose memory is being retrieved) and motor action reminiscent of arrival of that item/event. Work by Abdou et al., has two implicit assumptions. One is that engram cells interconnect between memories and secondly, synapse-specific plasticity ensures the identity and storage of individual memories. The findings in this work need mechanistic explanations for both behavioral motor action (withdrawal of foot) and, if possible, for inner sensation of memory of foot shock. Following is an explanation of how the findings in this paper can be explained in terms of semblance hypothesis. Alternatively speaking, findings in this paper allow us to look at semblance hypothesis from a new angle.

In fear conditioning experiments, two stimuli are associated. Out of this one (foot shock) generates a motor response (foot withdrawal). The other one has no motor response. The one with motor response is called unconditioned stimulus (US). The one that does not trigger any motor response on its own is called conditioned stimulus (CS). Authors associated between US and CS. After learning when CS arrives, motor action in response to the US (that occurred prior to learning) takes place.

At this juncture, we would like to find an explicit answer to the questions a) where associative learning is getting stored and how does memories get retrieved, and b) how behavioral motor activity reminiscent of arrival of a stimulus whose memory is being retrieved. We should understand them with such a clarity that we can explain them to an engineer who wants to replicate the mechanism in an engineered system. For this, the issue can be simplified as given in figure 1.

CS arrives through the auditory cortex (AC) & medial geniculate body. Medial division of the medial geniculate body (MGm or MGN) receives both auditory and somatosensory inputs (LeDoux et al., 1987; Bordi and LeDoux, 1994) and project to LA (LeDoux et al., 1990). Furthermore, it was verified that the US (foot shock) activates lateral amygdala neurons (Lanuza et al., 2008). Medical geniculae body is only a relay station for both auditory and foot shock stimuli without any interactions between them at this level. So, we need to imagine that AC-LA pathway is the auditory stimulus pathway and MGm-LA (MGN-LA) pathway is the footshock pathway. At the same time, we have to keep in mind that MGm (MGN) is the path through which AC connects to the LA. LA neurons receive inputs from both AC (sound) and MGm (foot shock). All the diagrams here refer to MGm path as MGN.

Figure 1. Conditioned stimulus (CS) arrives through neurons of the auditory cortex (ACN). Unconditioned stimulus (US) arrives through neurons of the medical geniculate nucleus (MGN). They synapse with different spines on a lateral amygdala (LA) neuron. All three of the above neurons are referred to as engram neurons. The question is "After associative learning, when the CS alone arrives, how does it trigger motor response (foot withdrawal) as if it is receiving a foot shock?" To answer this, it is necessary to show at least some evidence for an interaction between the spines of LA that synapse with stimuli arriving through both ACN and MGN in the figure.

The first question is, “What is the mechanistic explanation for the firing of lateral amygdala (LA) neuron when US comes after the associative learning between US and CS?” Authors provide and implicit explanation that plasticity changes occur at the spines of LA neuron on which inputs from CS and US synapse. A mechanistic explanation needs to meet 2 requirements. 1) How does arrival of CS alone cause firing of LA neuron reminiscent of arrival of US? 2) How does arrival of CS generate an internal sensation of arrival of US? If a single explanation can provide answers to both these two questions, then there is a good chance that it can be found correct after verification. At this time, readers can have more questions. First, why can't the inputs from two associatively learned item be shown synapsing to two neighboring spines on one LA neuron as follows (Figure 2)?

Figure 2. Inputs from two associatively learned stimuli arrive and synapse on to two neighboring spines on one LA neuron. Can this explain associative learning mechanism?

Looking at figure 2, we can ask the question, "Can it provide an explanation?" An interpretation of clustered plasticity along with synaptic tagging was explained previously by Govindarajan et al., 2006. The problems with this model are 1) there is no evidence for the formation of an electric cable between two neighboring spines, 2) there are no evidence for the formation of such a cable through the extracellular matrix space, 3) no evidence for the formation of specific tag molecules that can form in physiological timescales of milliseconds to explain function.

So, let us examine the underlying issue more closely by asking some questions.

1) How does arrival of CS cause firing of LA neuron reminiscent of arrival of US? An engineer will want to see at least the formation of a cable between the spines to which CS (ACN neuron) and US (MGN neuron) synapses. But there is no evidence for an electrical cable between them. Synaptic tagging was put forward as a mechanism (Frey and Morris, 1998). But sufficient number of specific tag molecules that can form and act at physiological timescales are not found. The solution must provide a mechanistic explanation.

2) Since there is no evidence for a long-lasting direct interaction between two spines that receive CS and US on an LA neuron, the configuration given Figure 2 is not compatible with the actual mechanism of learning. So, is there an alternative?

3) We can also ask, "How does arrival of CS generate an internal sensation of memory of arrival of US?" An ideal solution for the first question is expected to answer this question as well. Currently, we are not searching for a mechanism that generates inner sensations of memory due to several reasons that we can only assume. But an engineer who wants to replicate the mechanism will need to see a blueprint that explains a mechanism for this.  

So, the question is, "How to move forward to provide an explanation for both behavioral motor action reminiscent of memory retrieval and the very process of memory retrieval itself?" Both these are intricately connected parts of a single mechanism. Hence, we need to find such a mechanism that will also allow us to explain how the brain can store very large number of associated memories that are associated with firing of neuronal ensembles. Results from the present paper demands that the mechanism should be able to explain how memories are stored apparently using the same neuron in such way that they can be distinguished from each other.

This needs a renewed approach. First, let us see how the dendritic tree is organized. Are the tree branches of a neuron similar to the tree branches of a tree in a forest? Tree branches of a tree in a forest usually don't overlap with each other so that the leaves can get maximum sunlight. But the dendritic arbor of neighboring neurons overlap intensely so that we cannot separate the arbor of one neuron from another one (Figure 3).

Figure 3. Left side: Trees in a forest. The arbors of even closely located trees do not overlap with each other. Basic reason is that more trees beyond a limit cannot grow in forests due to limited sunlight that they can get. Right side: Drawing by Ramon y Cajal taken from "Comparative study of the sensory areas of the human cortex" page 363 showing Golgi-stained cortex of a 1.5-month-old infant. Cajal once referred to them as "impenetrable jungle". Here, we can see overlapping and intermingling of dendritic arbors of neighboring neurons. If we take a picture of the top view of the cortex to get a comparable view (like that of the forest), we won't be able to recognize any single neuronal arbor even if we paint each neuron differently. Note that Golgi staining stains only a small fraction of neurons! So, in reality, it is an "impenetrable jungle" (Figure on the left side is taken from Wikipedia). To get a clear picture of the functional importance of overlapping dendritic arbors, we must ask, "Spines of one neuron abut with spines of how many other neurons?"

What does this indicate? This means that there can be some functional significance for this. Also, dendritic branches of neurons from different cortical layers overlap with each other since their apical tuft regions stay attached to the inner surface of the pia mater, which is the covering of the cortex. This happens due to the descent of neurons from the sub-pial region towards the ventricle during development. I have explained this in figure 1 in a paper (Vadakkan, 2016). The advantage for such an organization is that a spine from one neuron can be present in between spines of a second neuron. Now we can ask, "Is there a possibility for interactions between neuronal processes of different neurons?" Specifically, since spines are the projecting parts on the dendrites of a neuron, we can ask, "Is there a possibility for the spines of a neuron to interact with neuronal process of other neurons?" For the occurrence of some physical interaction between the spine of one neuron and one of the neuronal processes of another neuron, it is reasonable to expect some space between the neighboring spines on the dendrite, even though in a 3D space this is not a necessity. So, we can ask, "Is the inter-spine distance greater than spine diameter?". Yes, it was found that mean inter-spine distance is more than mean spine diameter (Konur et al., 2003).

Remember, the US is a stimulus that will evoke foot withdrawal by itself. CS is a stimulus that does not evoke foot withdrawal. So naturally, CS won't synapse to an LA neuron that is connected to lower motor neurons to cause foot withdrawal. Hence, a configuration where inputs from both CS and US synapsing to the same LA neuron may appear to beat the very purpose of conditioned learning. So, for now we must change the configuration/components of all the above figures (Figures 1, 2 & 3) such that the LA neuron that receives input from CS is a different neuron than the LA neuron that receives input from US for demonstration purpose. Once everything becomes explainable, we will come back and ask this question again. So, for now inputs arriving towards LA neurons should be as in Figure 4. We can also assume that in the background state, LAN1 neuron is being held at a sub-threshold activated state that will allow sufficient potentials from MGN neuron (by the foot shock) to fire it. Let us see if this works.

Figure 4. In this configuration, inputs from associatively learned stimuli arrive to two different LA neurons through AC and MG neurons, so that only US stimulus can result in foot withdrawal before learning. According to this configuration, LA neurons that receive inputs from AC alone will not fire those LA neurons so that there won't be any foot withdrawal.

Now, if the spines of one LA neuron are located in the inter-spine spaces of the other LA neurons, then an inter-neuronal inter-spine interaction called inter-postsynaptic functional LINK (IPL) is formed between spines that belong to two different lateral amygdala (LA) neurons during learning. After learning, when the CS arrives, it can depolarize the spine on the second LA neuron that propagated stimuli from the US before learning and cause foot withdrawal (Figure 5). This can explain how motor action takes place reminiscent of the arrival of US even though US does not arrive at that time. Furthermore, reactivation of IPL by CS will depolarize the inter-LINKed spine of LAN1 in the absence of arrival of any stimulus from US. This will generate units of inner sensations (semblance) on the inter-LINKed spine of LAN1. For details, see FAQ section of this site.

Figure 5. Formation of inter-neuronal inter-spine IPL. Here, output from only LAN1 causes foot withdrawal. After learning, arrival of CS to LAN2 results in propagation of depolarization of the spine of LAN2 through the IPL to the spine of LAN1 to which US arrived at the time of learning. This will cause firing of sub-threshold activated neuron LAN1 and cause foot withdrawal reminiscent of arrival of US. Also, the reactivation of spine of LAN1 through the IPL will result in inner sensation of arrival of US (even though US does not arrive). For details, see FAQ section. The fact that a) dendritic arbors of neighboring neurons overlap with each other, and b) mean inter-spine distance is more than mean spine diameter (Konur et al., 2003), provide an opportunity for the presence of abutted spines that belong to different LA neurons. If associative learning can remove hydration layer between the outer layer of neuronal cell membranes over their spines, then it can lead to an electrical connection between them. When CS arrives, potentials will propagate through this connection and reach the LA (which is being maintained at a sub-threshold level) that can fire and result in a motor action (foot withdrawal) reminiscent of the arrival of US. For the duration of presence of inter-neuronal inter-spine connection, conditioned reflex can be elicited. The same electrical connection is referred to as inter-postsynaptic functional LINK (IPL) by semblance hypothesis. Reactivation of IPL by CS can explain generation of first-person inner sensation of memory of US (please see FAQ section of this website for details). ACN = Auditory cortical neuron. MGN = Medial geniculate neuron. LAN = Lateral amygdala neuron.

Now, one may ask a question, "Are there LA neurons with inputs only from CS (ACNs)?". The answer is "Not necessarily". LA neurons can have inputs from both CS and US. So contrary to what was said in the paragraph before figure 4, how does this work? Two conditions need to be satisfied. Before learning when CS arrives, LA neurons should not fire to generate foot withdrawal. Before learning when US arrives, LA neurons should fire to generate foot withdrawal. So, we can think of a configuration where an LA neuron receives inputs from both CS and US. The inputs from US provide more inputs (potentials) to allow the LA neuron to cross the threshold for firing; whereas inputs from CS provide very minimal inputs that will not allow LA neurons to fire. During learning, IPLs are formed between spines (that receive inputs from CS and US) that belong to different LA neurons. This will lead to the following change after learning - i. e. when CS arrives after learning, the stimuli will propagate through the IPLs formed during learning to generate foot withdrawal. It is reasonable to assume that after learning enough inputs will arrive from CS through several IPLs to one LA neuron that is held at sub-threshold activation state to cause its firing and lead to foot withdrawal. So figure 5 can be modified as figure 6.

Effect of inducing LTP & LTD

A mechanistic explanation for LTP occlusion experimental results in the paper can be explained in terms of the explanation of LTP based on semblance hypothesis (Vadakkan, 2019). This is as follows. Based on semblance hypothesis, both addition and removal of vesicles at the lateral borders of spines can increase and decrease membrane length respectively at these regions has an important consequence in the formation and reversal of IPLs. AMPA receptor exocytosis takes place at the lateral borders of spine head close to the synapse, a location where IPL formation is expected to take place.

In normal conditions, GluR1subunits of AMPA receptors are located on the spine membrane up to 25nm away from the synaptic junction (Jacob and Weinberg, 2015). This indicates a high probability that vesicles containing AMPAR subunits get exocytosed at this location and along with providing AMPAR subunits to form functional AMPARs, these vesicles increase spine surface areas at their lateral margins that can favor the formation of different types of IPLs. Studies have shown synaptic addition of GluR1 subunit–containing AMPA receptors during experimental induction of long-term potentiation (LTP) (Hayashi et al., 2000; Passafaro et al., 2001).

Since removal of membranes from lateral borders of spines is expected to occur during formation of vesicles for endocytosis, any endocytosis of AMPA receptors can reverse the IPLs. It is known that modest depolarization used in inducing long-term depression (LTD) causes AMPAR endocytosis (Lüscher and Malenka, 2012). Furthermore, it is known that modest depolarization by LTD stimulation protocols cause AMPA receptor endocytosis following activation of phosphatases that dephosphorylate AMPA receptors (Lüscher and Malenka, 2012). It was demonstrated that surface AMPARs are removed during induction of both NMDAR-dependent LTD (Beattie et al., 2000; Carroll et al., 2001), and mGluR-LTD (Waung et al., 2008; Park et al., 2008).

Fear conditioning generates large number of IPLs. So, between the stimulating and recording electrodes there are a large number of IPLs. Hence, LTD stimulation is expected to reverse some of these IPLs located between the electrodes. Since endocytosis of the AMPAR subunits and reduction in the size of enlarged spines can explain LTP decay (Dong et al., 2015), LTD can be explained in terms of reversal of existing IPLs in the brain tissue formed as a result of fear conditioning.

Effect of inducing autophagy

Effect of tat-beclin in erasing a memory can be explained as follows. Induction of autophagy by peptide tat-beclin leads autophagosome to fuse with endosome-lysosome system and degrades contents of the latter. This will lead to degradation of endosomes including those that contain AMPA receptors (glutamate receptor subtype with fast kinetics). When endosomes are formed and degraded continuously, IPLs will reverse back to form independent spines. This will result in failure to induce semblances when the cue stimulus (CS) arrives. Without further specific associative learning, these specific sets of IPLs will not be formed. This can explain how tat-beclin cause long-lasting memory erasure.

Independent storage of two memories in a shared ensemble of neurons

Authors used two types of CS (CS1 & CS2) to associate with one type of US. They have shown that erasing the association between one type of CS and US does not affect the association between second CS and US. Based on the semblance hypothesis, associations between different types of CS and US result in formation of specific IPLs. Autophagy results in memory destabilization and erasure of auditory fear memory associated with AMPAR endocytosis (Shehata et al., 2018). This can also lead to loss of membrane segments for endocytosis from the lateral borders of spines since GluR1subunits of AMPA receptors are located on the spine membrane up to 25 nm away from the synaptic junction (Jacob and Weinberg, 2015). This can lead to reversal of IPL formed with another spine of a second neuron. Reactivation of specific IPLs by a specific CS makes those IPLs vulnerable to reversal by endocytosis using membrane segments from lateral borders of the spines (how this occur in contrast to a resting IPL need to be studied) along with autophagy of the endocytosed vesicles containing AMPA receptor subunits.

Functional connectivity between engram cells (LA neurons). Where is it taking place?

Authors suggest the occurrence of a "functional connectivity" between engram cells. What contributes to this functional connectivity? How does the brain store enormous number of specific memories, retrieved in response to specific cue stimulus, in shared cell assemblies? Explanations of the findings in the paper by Abdou et al., were made based on the views that memories are formed by changes in synaptic plasticity and are stored in specific neuronal ensembles. Semblance hypothesis used inductive reasoning approaches to derive a mechanism that takes place during learning that has the potential for generating both behavioral motor actions and induce first-person inner sensation of memory. Based on this, inter-neuronal inter-spine functional LINKs are formed during learning and are reactivated during memory retrieval providing the functions. This mechanism operates in synchrony with the normal synaptic functions and neuronal firing.

Based on the semblance hypothesis, specific IPLs are formed between spines that belong to specific neurons during a specific learning. Depending on the specific features of a cue stimulus, a set of IPLs get reactivated. Those IPL reactivations that contribute to potentials allowing subthreshold neurons to cross the threshold will fire. These firing cells are called "engram cells". Along with this, IPL reactivations generate inner sensations (semblance) of memory of the associatively learned item. This takes place as a system property of systems where synaptic transmission and propagation of potentials across the IPL contribute vector components to the oscillating extracellular potentials. Note that the brain functions only during a narrow range of frequency of oscillating extracellular potentials. Thus, IPLs provide what the authors refer to as "functional connectivity" between assemblies of firing cells. Since large number of different IPL reactivations can independently contribute potentials to neurons that are being held at subthreshold states, it will provide an apparent view that large number of memories are stored using shared cell assemblies. Explanations provided here constitute a retrodictive (postdictive) evidence in support of the semblance hypothesis.

Beattie EC, Carroll RC, Yu X, Morishita W, Yasuda H, von Zastrow M, Malenka RC (2000) Regulation of AMPA receptor endocytosis by a signaling mechanism shared with LTD, Nat. Neurosci. 312 (2000) 1291–1300. PubMed

Bordi F, LeDoux JE (1994) Response properties of single units in areas of rat auditory thalamus that project to the amygdala. II. Cells receiving convergent auditory and somatosensory inputs and cells antidromically activated by amygdala stimulation. Exp Brain Res. 98:275–286. PubMed

Carroll RC, Beattie EC, von Zastrow M, Malenka RC (2001) Role of AMPA receptor endocytosis in synaptic plasticity, Nat. Rev. Neurosci. 315–324. PubMed

Dong Z, Han H, Li H, Bai Y, Wang W, Tu M, Peng Y, Zhou L, He W, Wu X, Tan T, Liu M, Wu X, Zhou W, Jin W, Zhang S, Sacktor TC, Li T, Song W, Wang YT (2015) Long-term potentiation decay and memory loss are mediated by AMPAR endocytosis. J Clin Invest. 125(1):234-247. PubMed

Jacob AL, Weinberg RJ (2015) The organization of AMPA receptor subunits at the postsynaptic membrane, Hippocampus 25:798–812. PubMed

Frey U, Morris RG (1998) Synaptic tagging: implications for late maintenance of hippocampal long-term potentiation. Trends Neurosci. 21(5):181-188. PubMed

Govindarajan A, Kelleher RJ, Tonegawa S (2006) A clustered plasticity model of long-term memory engrams. Nat Rev Neurosci. 7(7):575-583. PubMed

Hayashi Y, Shi SH, Esteban JA, Piccini A, Poncer JC, Malinow R (2000) Driving AMPA receptors into synapses by LTP and CaMKII: requirement for GluR1 and PDZ domain interaction. PubMed

Konur S, Rabinowitz D, Fenstermaker VL, Yuste R (2003) Systematic regulation of spine sizes and densities in pyramidal neurons. J Neurobiol. 56(2):95-112. PubMed

Lanuza E, Moncho-Bogani J, Ledoux JE (2008) Unconditioned stimulus pathways to the amygdala: effects of lesions of the posterior intralaminar thalamus on foot-shock-induced c-Fos expression in the subdivisions of the lateral amygdala. Neuroscience. 155(3):959-956. PubMed

Ledoux JE, Ruggiero DA, Forest R, Stornetta R, Reis DJ (1987) Topographic organization of convergent projections to the thalamus from the inferior colliculus and spinal cord in the rat. J Comp Neurol. 264:123–146. PubMed

LeDoux JE, Farb C, Ruggiero DA (1990) Topographic organization of neurons in the acoustic thalamus that project to the amygdala. J Neurosci. 10:1043–1054. PubMed

Lüscher C, Malenka RC (2012) NMDA receptor-dependent long-term potentiation and long–term depression LTP/LTD, Cold Spring Harb. Perspect. Biol. 4 (6) a005710. PubMed

Park S, Park JM, Kim S, Kim JA, Shepherd JD, Smith-Hicks CL, Chowdhury S, Kaufmann W, Kuhl D, Ryazanov AG, Huganir RL, Linden DJ, Worley PF (2008) Elongation factor 2 and fragile X mental retardation protein control the dynamic translation of Arc/Arg3.1 essential for mGluR-LTD, Neuron 59 (1):70–83. PubMed

Passafaro M, Piëch V, Sheng M (2001) Subunit-specific temporal and spatial patterns of AMPA receptor exocytosis in hippocampal neurons PubMed

Shehata M, Abdou K, Choko K, Matsuo M, Nishizono H, Inokuchi K (2018) Autophagy Enhances Memory Erasure through Synaptic Destabilization. J Neurosci. 38(15):3809-3822. PubMed

Vadakkan KI (2016) Rapid chain generation of interpostsynaptic functional LINKs can trigger seizure generation: Evidence for potential interconnections from pathology to behavior. Epilepsy Behav. 59:28-41. PubMed

Vadakkan KI (2019) A potential mechanism for first-person internal sensation of memory provides evidence for the relationship between learning and LTP induction. Behav Brain Res. 360:16-35. PubMed

Waung MW, Pfeiffer BE, Nosyreva ED, Ronesi JA, Huber KM (2008) Rapid translation of Arc/Arg3.1 selectively mediates mGluR-dependent LTD through persistent increases in AMPAR endocytosis rate, Neuron 591: 84–97. PubMed

Evidence for long-term potentiation in phospholipid membranes. Scott HL, Bolmatov D, Podar PT, Liu Z, Kinnun JJ, Doughty B, Lydic R, Sacci RL, Collier CP, Katsaras J (2022 December 13^th) Proc Natl Acad Sci U S A. 119(50):e2212195119. PubMed

Long-term potentiation (LTP) is an electrophysiological finding observed in large number of brain regions where different inputs converge. Large number of correlations were found between LTP and the ability of animals to store associatively learned information and retrieve their memories. However, it is necessary to verify whether LTP can generate cellular changes similar to that occur in physiological conditions in vivo.

It was shown that the least path required to induce LTP is the stimulation of postsynaptic terminals (dendritic spines) (Kullmann et al. 1992). Later, it was shown that inhibitors that block postsynaptic membrane fusion attenuate LTP (Lledo et al., 1998) even though it is not clear where exactly this fusion takes place. Using a derived mechanism for generating first-person inner sensations (that provides testable predictions), it was possible to explain LTP in terms of inter-neuronal inter-spine membrane interactions that were called inter-postsynaptic functional LINKs (IPLs). Structure of IPLs varies depending on the inter-membrane interaction that can range between exclusion of hydration layer between membranes and membrane hemifusion (Vadakkan, 2019a). The delay for induction of LTP was explained in terms of time taken for the dendritic spines in a field (containing large number of synapses at a location between stimulating and recording electrodes) to expand and start interacting (Vadakkan, 2019a) (video). Testable predictions for the operational mechanism of IPLs were provided (Vadakkan, 2019b).

Before undertaking any in vivo experiments, it is necessary to obtain evidence to suggest whether inter-spine membrane interactions can cause a long-lasting electrical connection between them. Semblance hypothesis proposed displacement of hydration water, from the extracellular space between spine membranes of different neurons at the location of their convergence, that can progress up to the formation of hemifusion between those spine membranes. Semblance hypothesis also proposed that the area of membrane interaction can act as a passive conductive location between spines that belong to different neurons. Furthermore, this interaction enables input signals arriving from one of the associatively learned stimuli to depolarize the interacting second membrane to generate inner sensation of the second stimulus due to certain specific features at the synapses (explained in Vadakkan, 2013).

A new evidence suggests that inter-membrane interactions between lipid monolayers can lead to membrane hemifusion and generate capacitance at those locations where energy can be stored (Scott et al., 2022). Slow discharge from this capacitor enables these hemifused locations to produce a long-lasting potentiated effect similar to that of LTP. The ability to store charge and its discharge is subsequently match with a property having time-scales suitable to explain working memory. But storing and releasing charge can only explain the electrical findings. More questions remain. "Can such a property be explained between vesicles whose membranes are lipid bilayers?", “How does such inter-membrane interaction contribute to inner sensation of memories?” Semblance hypothesis has provided an explanation of how semblances can be induced at the second inter-LINKed spine as a system property when one of the associatively learned stimuli arrive at the first inter-LINKed spine (Vadakkan, 2013, 2019).

The findings in Scott et al’s work show that inter-membrane interaction has the ability to retain hemifused area for certain length of time. If such changes can occur in biological system, it will enable the neuronal cells to take steps to stabilize those interactive areas for a long period. If two associated stimuli continue to arrive frequently, then it is better for the neuronal cells to stabilize the inter-neuronal inter-spine interactive region, if it provides certain cellular level benefits. While this is the primary aim as far as the cells are concerned, it leaves the system with a functional advantage to generate semblances for inner sensations that will enable the animal to survive, which can be viewed as an evolutionarily gained feature.

Scott et al’s work provides an explanation that the hemifusion between lipid monolayers can remain stable for certain period without causing mixing of the vesicle contents. It also indicates the possibility that hemifusion between double membranes can be undertaken in future studies to examine for such possibilities. How does capacitance development found by Scott et al at the locations of hemifusion influence the hemifused areas between lipid bilayers of spines that belong to different neurons? Due to the reason that mean inter-spine distance is more than mean spine head diameter (Konur et al., 2003), inter-neuronal inter-spine interaction leading to IPL formation has been viewed as a candidate mechanism.

Biological systems have several trans-membrane proteins whose extracellular domains can interact in such a way to exclude hydration layer between membranes to promote inter-membrane interactions. For example, SNARE (soluble NSF (N-ethylmaleimide sensitive fusion protein) attachment protein receptor) proteins are known to provide energy for bringing together membranes against repulsive charges and overcome energy barrier related to curvature deformations during hemifusion between abutted membranes (Martens and McMahon, 2008; Oelkers et al., 2016). They also generate force to pull together abutted membranes as tightly as possible (Hernandez et al., 2012). By initiating the fusion process by supplying energy (Jahn and Scheller, 2006), SNARE proteins can lead to the formation of characteristic hemifusion intermediates (Lu et al., 2005; Liu et al., 2008). These properties of SNARE proteins highlight their functional significance to form hemifusion intermediates that can possibly contribute to IPL formation. These intracellular factors can contribute to a short-lasting inter-spine membrane interaction leading to the generation of a short-lived IPL.

There are several avenues for futures studies in lipid chemistry to understand inter-membrane interactions at locations in the nervous system where extracellular matrix space is very minimal. Manipulating membrane fusion stages using SNARE modulators (Khvotchev and Soloviev, 2022) to understand its effects on LTP, associative learning and memory can lead to further advancements in this area of research.

Hernandez JM, Stein A, Behrmann E, Riedel D, Cypionka A, Farsi Z, Walla PJ, Raunser S, Jahn R (2012) Membrane fusion intermediates via directional and full assembly of the SNARE complex. Science 336(6088):1581–84 PubMed

Jahn R, Scheller RH (2006) SNAREs--engines for membrane fusion. Nat Rev Mol Cell Biol 7(9):631–43 PubMed

Khvotchev M, Soloviev M (2022) SNARE Modulators and SNARE Mimetic Peptides. Biomolecules. 12(12):1779. PubMed

Konur S, Rabinowitz D, Fenstermaker VL, Yuste R (2003) Systematic regulation of spine sizes and densities in pyramidal neurons. J Neurobiol. 56(2):95-112. PubMed

Kullmann DM, Perkel DJ, Manabe T, Nicoll RA (1992) Ca2+ entry via postsynaptic voltage-sensitive Ca2+ channels can transiently potentiate excitatory synaptic transmission in the hippocampus. Neuron. 9(6):1175-83. Pubmed

Liu T, Wang T, Chapman ER, Weisshaar JC. Productive hemifusion intermediates in fast vesicle fusion driven by neuronal SNAREs. Biophys J 94(4):1303–14 (2008). PubMed

Lledo PM, Zhang X, Südhof TC, Malenka RC, Nicoll RA (1998) Postsynaptic membrane fusion and long-term potentiation. Science. 279(5349):399-403. PubMed

Lu X, Zhang F, McNew JA, Shin YK (2005) Membrane fusion induced by neuronal SNAREs transits through hemifusion. J Biol Chem 280(34):30538–41. PubMed

Martens S, McMahon HT (2008) Mechanisms of membrane fusion: disparate players and common principles. Nat Rev Mol Cell Biol 9(7):543–556 PubMed

Oelkers M, Witt H, Halder P, Jahn R, Janshoff A (2016) SNARE-mediated membrane fusion trajectories derived from force-clamp experiments. Proc Natl Acad Sci U S A 113(46):13051–56. PubMed

Scott HL, Bolmatov D, Podar PT, Liu Z, Kinnun JJ, Doughty B, Lydic R, Sacci RL, Collier CP, Katsaras J (2022) Evidence for long-term potentiation in phospholipid membranes. Proc Natl Acad Sci U S A. 119(50):e2212195119. PubMed

Vadakkan KI (2013) A supplementary circuit rule-set for the neuronal wiring. Front Hum Neurosci. 7:170. PubMed

Vadakkan KI (2019a) A potential mechanism for first-person internal sensation of memory provides evidence for the relationship between learning and LTP induction. Behav Brain Res. 360:16-35. PubMed

Vadakkan KI (2019b) From cells to sensations: A window to the physics of mind. Phys Life Rev. 31:44-78. PubMed

Entorhinal cortex directs learning-related changes in CA1 representations Grienberger and Magee, November 2^nd 2022. Nature. doi: 10.1038/s41586-022-05378-6 PubMed

Rapid synaptic plasticity contributes to a learned conjuctive code of position and choice-related information in the hippocampus (Zhao X, Hsu CL, Spruston N, January 5, 2022) Neuron. 110(1):96-108.e4. PubMed

It is known that Entorhinal cortex (EC) input to hippocampus (HP) is via two different projections. 1) A trisynaptic path from EC layer II (ECII) to CA1 neurons via CA3 neurons and a monosynaptic pathway directly connecting EC layer III (ECIII) to CA1 neurons (Fig.1).

Figure 1. Figure showing both trisynaptic and monosynaptic pathways between one entorhinal cortical (EC) neurons and one CA1 pyramidal neuron. Trisynaptic path from EC2 neurons connects through granule neurons and neurons of the CA3 layer. Monosynaptic path from EC3 neurons synapse to a CA1 neuron located at stratum lacunosum-moleculare layer at apical tuft region of CA1 neuron. At the very least, it is necessary to explain how inner sensation of memory of location is generated and how it is related to firing of a set of CA1 neurons in the presence of a cue stimulus. Note: Colored circles are neurons.

CA1 pyramidal neurons that fire somatic action potentials when the animal reaches a specific location are called place cells (Moser et al., 2015). Results from one study has led to the inference that EC3 to CA1 connections are involved in temporal association memory (Suh et al, 2010). Later, it was noticed that ability to associatively learn and memorize a location in response to a cue stimulus is correlated with increased firing of CA1 neurons (Zhao et al., 2020) as if there is an “over-representation” of these neurons to place memory. It was also noticed that firing of CA1 neurons is associated with long-term dendritic voltage signals initiated by inputs from EC3 sub-domain of EC (Magee and Grienberger, 2020). Authors attribute this to occurrence of behavioral timescale synaptic plasticity (BTSP) in the EC3-CA1 synapses. Recent experiments that showed increased elevation of both EC3 activity and CA1 place field density in response to a prominent reward-predictive cue stimulus in a new environment led to the interpretation that EC directs learning-related changes in CA1 representations (Grienberger and Magee, 2022). Similar results reported earlier by Zhao et al., 2022 inferred that plateau potentials in CA1 pyramidal neurons rapidly strengthen synaptic inputs carrying conjunctive information about position and choice. Investigators also inferred that learning is due to the formation of a conjunctive population code upstream of CA1. However, a mechanistic explanation with a rigor for replicating the mechanism in an engineered system is not available.

To provide a mechanistic explanation for the above findings, it is necessary to arrive at a mechanism that can provide explanations for the following questions (Table 1).

Table 1. Questions that a solution for the nervous system is expected to provide answers for in an interconnected manner.

Challenges and features of a possible solution

Work by Grienberger and Magee (Grienberger and Magee, 2022) infers that there are synaptic plasticity changes at the EC3-CA1 synapses that lead to dendritic voltage signals in the dendrites of CA1 neurons, which is associated with/in turn leads to firing of CA1 neurons. Increased firing of CA1 neurons is being interpreted as “overrepresentation”. A mechanistic explanation is necessary to explain both “plasticity” and “over-representation” with the type of clarity that will allow its replication in engineered systems. The above requirements given in Table 1 can be summarized to two questions. What synaptic changes inferred from dendritic voltage signals, which are being referred to as synaptic plasticity changes, can generate first-person inner sensation of memory? How is the same mechanism linked to sudden firing of previously silent CA1 neurons that made us to infer that they acquire place field property?

The ability to retrieve memory is currently being studied using surrogate markers such as behavioral motor actions and speech. Instead of examining surrogate markers, semblance hypothesis searched for a mechanism for first-person inner sensations directly by asking the question, “At what location and by what mechanism do the first-person inner sensations get sparked?” Even though third person experimenter cannot sense or identify the formation for first-person properties at this location, reaching a solution point provides a mechanistic explanation for the most important function of the nervous system. It can lead to finding methods to treat its disorders and replicating the mechanism in engineered systems. This motivated derivation of semblance hypothesis (Vadakkan, 2007, 2013, 2019b). It was based on the argument that if it becomes possible to formulate a mechanism for generating inner sensations that can also explain all features of the system exhibited in different levels by an interconnectable mechanism, then the formulated mechanism can be correct. It will then be possible to make testable predictions that can be verified.

Explanation

Towards achieving this, first a conditional definition for memory was made (Vadakkan, 2007). This was followed by examining works that were carried out with an aim to undertake the gold standard test of replicating the mechanism in engineered systems, which will eventually lead to the development of true artificial intelligence (AI). A pioneering work (Minsky, 1980) that viewed memories as hallucinations (internal sensations of something in its absence) matched with the expectations of search for a mechanism of first-person inner sensations. This laid a foundational framework for a testable mechanism. Motivated by this, a search was carried out in the nervous system to identify a change that can occur during associative learning and can be used by one of the associatively learned stimuli to generate hallucinations of second stimulus at the time of memory retrieval.

In the background state, head region of a dendritic spine (postsynaptic or input terminal) is continuously getting depolarized by quantal release of neurotransmitter molecules, in addition to occasional volleys of release of neurotransmitter molecules when action potentials arrive at its presynaptic terminal. Simultaneous activation of two abutted spines by environmental stimuli is expected to form inter-postsynaptic (inter-spine) functional LINK (IPL) during associative learning (Vadakkan, 2013), which forms the linchpin of derived mechanism. At the time of memory retrieval, reactivation of this IPL by one of the associatively learned stimuli leads to propagation of potentials to the inter-LINKed spine previously activated by the second stimulus whose memory is expected to get retrieved. In the background state of continuous depolarization of spine head by quantal release of neurotransmitter molecules, any sudden lateral activation of inter-LINKed spine (in the absence of arrival of action potentials at its presynaptic terminal) is expected to generate a hallucination that the inter-LINKed spine is receiving a stimulus from the environment through its presynaptic terminal. Qualia of first-person inner sensations of a retrieved memory can be estimated by retrograde extrapolation from the inter-LINKed spine towards identifying all the sensory receptors (see figures 6 and 7 in FAQ section of this website). A unit of semblance (semblion) is equivalent to minimum sensory stimuli capable of stimulating a minimum subset of sensory receptors that will stimulate the inter-LINKed spine. A natural retrograde extrapolation is expected to occur at the time of memory retrieval as a system property of systems where synaptic transmission and propagation of potentials across the IPLs contribute intracellular potentials, whose corresponding changes in the extracellular matrix (ECM) space form vector components of oscillating extracellular potentials taking place within in a narrow range of frequencies. Answers to questions in Table1 are given in Table 2.

Table 2. Answers to questions that a solution for the nervous system is expected to provide answers for in an interconnected manner. The IPL mechanism has provided interconnected explanations for large number of findings in the system and has generated several testable predictions (Vadakkan, 2019b).

Even though semblions from stratum lacunosum-moleculare and stratum radiatum layers appear to be occurring at two different locations, large extracellular waveforms such as theta that connects both these locations (see Fig.1 in Fernández-Ruiz et al., 2017) informs about a mechanism for integrating (binding) semblions from these two locations. Note: Small colored circles are dendritic spines participating in inter-postsynaptic functional LINKs (IPLs). Note: Colored circles are dendritic spines (postsynaptic terminals).

Figure 2. Figure showing how trisynaptic and monosynaptic pathways from entorhinal cortical (EC) neurons and a CA1 pyramidal neuron form islets of inter-LINKed spines in the stratum lacunosum-moleculare and stratum radiatum layers respectively. In this demonstration, only five CA1 neurons with one of their spines participating in the islets of inter-LINKed spines is shown. Also, inter-LINKed spines in one islet are expected to belong to different CA1 pyramidal neurons as indicated by different colors (This can vary. For example, in dendritic excrescences of CA3 neurons). Since dendritic arbor at the apical tuft region where EC3 direct input arrives is relatively bigger, it is reasonable to expect formation of large number of IPLs at this location. This may explain an increased horizontal component contributing to low frequency theta wavefsorms at this location. Even though semblions from stratum lacunosum-moleculare and stratum radiatum layers appear to be occurring at two different locations, large extracellular waveforms such as theta that connects both these locations (see Fig.1 in Fernández-Ruiz et al., 2017) informs about a mechanism for integrating (binding) semblions from these two locations. Note: Small colored circles are dendritic spines participating in inter-postsynaptic functional LINKs (IPLs). Note: Colored circles are dendritic spines (postsynaptic terminals).

Oscillating extracellular potentials show characteristic waveforms in these two layers (Fernández-Ruiz et al., 2017) reflecting the nature of spines that are involved in forming IPLs in these locations. Both amplitude and frequency of these wave forms can be explained in terms of vector components contributed by IPLs at these locations. Low frequency theta waveforms can also be explained as the net effect of the vector components contributed by IPLs in a larger area. (see Fig.1 in https://www.cell.com/neuron/pdfExtended/S0896-6273(17)30101-0 and extracellular recordings: https://www.nature.com/articles/nn.2894/figures/1). Large number of interneurons are present in the L-M region (Capogna, 2011) indicates presence of IPLs between their spines and spines of CA1 neurons modifying the qualia of internal sensations generated at this location. The net potential generated at the islet of inter-LINKed spines propagates to the axon hillock of the CA1 neurons, allowing some of the sub-threshold activated CA1 neurons to fire somatic action potential.

Firing of EC3 neurons followed by firing of CA1 neurons prompt one to infer involvement of EC3-CA1 synaptic changes. Inhibition of NMDA receptor channels at these synapses causes inhibition of both learning and memory retrieval. The involvement of synapses towards the formation and reactivation of IPLs can be interpreted as synaptic plasticity changes involving EC3-CA1 synapses. Since IPL mechanism can explain generation of inner sensations can occur concurrent with firing of CA1 neurons, it can be viewed as a better explanation. Large amplitude synaptic inputs delivered by EC3 axons in apical dendritic tree of CA1 (Megias et al., 2001; Steward and Scoville, 1976) can lead to the formation of IPLs between spines of different CA1 neurons located in the stratum lacunosum-moleculare layer at apical tuft region of CA1 neuron. Both increased probability and duration of plateau potentials (Takahashi and Magee, 2009; Bittner et al., 2015) can be explained in terms of propagation of potentials through islets of inter-LINKed spines formed by large number of abutted spines stimulated. IPL reactivation provides additional potentials to sub-threshold activated CA1 neurons allowing them to fire an action potential.

Using CA1 neurons that fire, hippocampal maps were created to study their association for spatial memory performance (Dupppret et al., 2010), which led to the inference that accumulation of place fields is responsible for spatial learning. A pyramidal neuron that has thousands of input terminals can fire a somatic action potential when nearly 140 inputs signals arrive at any combination of input terminals (Eyal et al., 2018). Extreme degeneracy of input signals in firing a CA1 neuron makes firing of a CA1 neuron non-specific with respect to the location from where potentials arrive. Furthermore, large plateau potential anticipated to be generated by islet of inter-LINKed spines leads to more non-specificity of CA1 neuronal firing with respect to the input signals. Inhibition of CA1 firing by AP5 (antagonize NMDA receptors at the synapses of spines of CA1 neuron) or inhibitor of plateau firing CAV2-3 channel blocker SNX-482 occur due to inhibition of neurons. Since synapses are necessary for IPL mechanism to operate, chemicals that block synaptic functions will stop cognitive function and CA1 neuronal firing (place cell firing).

Conclusion

Many times, interpretations of experimental findings are carried out with the presupposition that a single neuron process information. What prompts such interpretations? A neuron having thousands of input terminals and receiving nearly any 140 input signals (less than this during temporal summation) can fire a somatic action potential (Eyal et al., 2018). Hence, the firing property of a neuron cannot be viewed as something that process information. But there is an information processing that must be occurring in the neighborhood of a neuron. Since there is extreme degeneracy of input signals in firing a neuron (Vadakkan, 2018), information processing mechanism is expected to take place in the vicinity of input terminals. It is necessary to ask basic questions to obtain a mechanistic explanation. The solution for the system should be able to explain how first-person property of inner sensation of a place occur in the nervous system along with remaining features such as 1) optional concurrent behavioral motor actions, and 2) firing of CA1 neurons. Interaction between spines of different neurons separated by narrow extracellular matrix provides a solution for the system. EC3 inputs synapse with spines of CA1 neuron in the stratum lacunosum-moleculare layer at the dendritic apical tuft region. Since spines of adjacent pyramidal neurons overlap with each other, interaction between these spines is expected to occur to generate inner sensations. When these interactions contribute to additional potentials to a subthreshold activated pyramidal neuron, it fires a somatic action potential, which is being viewed as place cells.

The observation that CA1 firing is inhibited by inhibitor of plateau firing Ca_v2.3 Ca²⁺channel blocker indicates that it is capable of blocking somewhere along the route from EC3 to CA1. Explanations for different phenomena indicated presence of IPLs between spines of different CA1 neurons. For example, Ca_v2.3 Ca²⁺ channels mediate epileptiform activity such as afterdepolarization, plateau potentials and exacerbation of low-threshold Ca²⁺ spikes resulting in seizure initiation and propagation (Wormuth et al., 2016). It was shown how IPL mechanism can explain seizure generation (Vadakkan, 2016). Ca_v2.3 Ca²⁺ channels in the presynaptic terminals are involved in LTP (Breustedt et al., 2003). It was previously explained how synapses are involved in IPL formation and reactivation to influence LTP (Vadakkan, 2019a). Furthermore, Ca_v2.3 Ca²⁺ channels have a highly organized spatial distribution with predominant expression in the proximal or distal dendrites (Westenbroek et al., 1995).

The IPL mechanism provided by semblance hypothesis can explain how “engram neurons” that are often seen as “representations” of memory and “synaptic plasticity” change. Conducting experiments to study interaction between abutted spines that belong to different neurons by the arrival of two associatively learned stimuli can be used to verify the presence of IPLs. GFP-reconstitution method across synaptic partners (GRASP) (Feinberg et al., 2008; Gordon and Scott, 2009; Fan et al., 2013; Macpherson et al., 2015; Shearin et al., 2018) was used to study synaptic connections. Trans-Tango (Talay et al., 2017) and TRACT (Huang et al., 2017) were invented for anterograde trans-synaptic tracing. Retrograde trans-synaptic tracing was carried out using a method called BAcTrace, (Cachero et al., 2020). Recently, retro-Tango method of retrograde synaptic tracing was developed (Sorkaç et al., 2022). Since TRACT, trans-Tango and retro-Tango methods allow neurons to show synaptic partners, similar approaches can be utilized to develop inter-spine tracers to study IPL links between spines that belong to different neurons of one neuronal order. Explanations provided here are expected to motivate verification of interactions between spines belonging to different neurons.

Bittner KC, Grienberger C, Vaidya SP, Milstein AD, Macklin JJ, Suh J, Tonegawa S, Magee JC (2015) Conjunctive input processing drives feature selectivity in hippocampal CA1 neurons. Nat. Neurosci. 18:1133–1142. PubMed

Breustedt J, Vogt KE, Miller RJ, Nicoll RA, Schmitz D (2013) Alpha1E-containing Ca2+ channels are involved in synaptic plasticity. Proc. Natl. Acad. Sci. USA. 100(21):12450–12455. PubMed

Cachero S, Gkantia M, Bates AS, Frechter S, Blackie L, McCarthy A, Sutcliffe B, Strano A, Aso Y, Jefferis G (2020) BAcTrace, a tool for retrograde tracing of neuronal circuits in Drosophila. Nat. Methods, 17(12):1254–1261. PubMed

Capogna M (2011) Neurogliaform cells and other interneurons of stratum lacunosum-moleculare gate entorhinal-hippocampal dialogue. J. Physiol. 589(Pt 8):1875–1883. PubMed

Eyal G, Verhoog MB, Testa-Silva G, Deitcher Y, Benavides-Piccione R, DeFelipe J, de Kock CPJ, Mansvelder HD, Segev I (2018) Human cortical pyramidal neurons: From spines to spikes via models. Front. Cell Neurosci.12:181. PubMed

Fan P, Manoli DS, Ahmed OM, Chen Y, Agarwal N, Kwong S, Cai AG, Neitz J, Renslo A, Baker BS, Shah NM (2013) Genetic and neural mechanisms that inhibit Drosophila from mating with other species. Cell. 154(1):89–102. PubMed

Feinberg EH, Vanhoven MK, Bendesky A, Wang G., Fetter RD, Shen K, Bargmann CI (2008) GFP reconstitution across synaptic partners (GRASP) defines cell contacts and synapses in living nervous systems. Neuron. 57(3):353–363. PubMed

Fernández-Ruiz A, Oliva A, Nagy GA, Maurer AP, Berényi A, Buzsáki G (2017) Entorhinal-CA3 dual-input control of spike timing in the hippocampus by theta-gamma coupling. Neuron. 93(5):1213–1226.e5. PubMed

Gordon MD, Scott K (2009) Motor control in a Drosophila taste circuit. Neuron. 61(3):373–384. PubMed

Grienberger C, Chen X, Konnerth A (2014) NMDA receptor-dependent multidendrite Ca2+ spikes required for hippocampal burst firing in vivo. Neuron. 81:1274–1281. PubMed

Hollup SA, Molden S, Donnett JG, Moser MB, Moser EI (2001) Accumulation of hippocampal place fields at the goal location in an annular watermaze task. J. Neurosci. 21:1635–1644. PubMed

Huang TH, Niesman P, Arasu D, Lee D, De La Cruz AL, Callejas A, Hong EJ, Lois C (2017) Tracing neuronal circuits in transgenic animals by transneuronal control of transcription (TRACT). Elife. 6.664. PubMed

Macpherso LJ, Zaharieva EE, Kearney PJ, Alpert MH, Lin TY, Turan Z, Lee CH, Gallio M (2015) Dynamic labelling of neural connections in multiple colours by trans-synaptic fluorescence complementation. Nat. Commun. 6:10024. PubMed

Magee JC, Grienberger C (2020) Synaptic plasticity forms and functions. Annu. Rev. Neurosci. 43:95–117. PubMed

Megias M, Emri Z, Freund TF, Gulyas AI (2001) Total number and distribution of inhibitory and excitatory synapses on hippocampal CA1 pyramidal cells. Neuroscience. 102:527–540. PubMed

Minsky M (1980) K-lines: a theory of memory. Cognitive Science 4:117–133 Article

Moser MB, Rowland DC, Moser EI (2015) Place cells, grid cells, and memory. Cold Spring Harb Perspect Biol. 7(2):a021808. PubMed

Shearin HK, Quinn CD, Mackin RD, Macdonald IS, Stowers RS (2018) t-GRASP, a targeted GRASP for assessing neuronal connectivity. J. Neurosci. Methods. 306:94–102. PubMed

Sorkaç A, Moșneanu RA, Crown AM, Doruk Savaş D, Okoro AM, Talay M, Barnearetro G (2022) retro-Tango enables versatile retrograde circuit tracing in Drosophila. PubMed Biorxiv

Steward O, Scoville SA (1976) Cells of origin of entorhinal cortical afferents to the hippocampus and fascia dentata of the rat. J. Comp. Neurol. 169:347–370. PubMed

Suh J, Rivest AJ, Nakashiba T, Tominaga T, Tonegawa S (2011) Entorhinal cortex layer III input to the hippocampus is crucial for temporal association memory. Science. 334:1415–1420. PubMed

Takahashi H, Magee JC (2009) Pathway interactions and synaptic plasticity in the dendritic tuft regions of CA1 pyramidal neurons. Neuron. 62:102–111. PubMed

Talay M, Richman EB, Snell NJ, Hartmann GG, Fisher JD, Sorkac A, Santoyo JF, Chou-Freed C, Nair N, Johnson M, Szymanski JR, Barnea G (2017) Transsynaptic mapping of second-order taste neurons in flies by trans-Tango. Neuron. 96(4):783–795 e784.762. PubMed

Vadakkan KI (2007) Semblance of activity at the shared post-synapses and extracellular matrices - A structure function hypothesis of memory. ISBN:978-0-5954-7002-0

Vadakkan KI (2010) Framework of consciousness from semblance of activity at functionally LINKed postsynaptic membranes. Front. Psychol. 1:168 PubMed

Vadakkan KI (2013) A supplementary circuit rule-set for the neuronal wiring. Frontiers in Human Neuroscience. 1;7:170. PubMed

Vadakkan KI (2015) A pressure-reversible cellular mechanism of general anesthetics capable of altering a possible mechanism for consciousness. Springerplus. 4: 485. PubMed

Vadakkan KI (2019a) A potential mechanism for first-person internal sensation of memory provides evidence for the relationship between learning and LTP induction. Behav Brain Res. 360:16-35. PubMed

Vadakkan KI (2019b) From cells to sensations: A window to the physics of mind. Phys Life Rev. 31:44-78. PubMed

Vadakkan KI (2018) Extreme degeneracy of inputs in firing a neuron leads to loss of information when neuronal firing is examined. PeerJ Preprints

Vadakkan KI (2020) A Derived Mechanism of Nervous System Functions explains Aging-Related Neurodegeneration as a Gradual Loss of an evolutionaly adaptation. Current Aging Science. 13(2): 136-152. PubMed

Westenbroek RE, Sakurai T, Elliott EM, Hell JW, Starr TV, Snutch TP, Catterall WA (1995) Immunochemical identification and subcellular distribution of the alpha 1A subunits of brain calcium channels. J. Neurosci. 15(10):6403–6418. PubMed

Wormuth C, Lundt A, Henseler C, Müller R, Broich K, Papazoglou A, Weiergräber M (2016) Review: Cav2.3 R-type voltage-gated Ca2+ channels - Functional implications in convulsive and non-convulsive seizure activity. Open Neurol J. 10:99-126. PubMed

Zhao X, Wang Y, Spruston N, Magee JC (2020) Membrane potential dynamics underlying context-dependent sensory responses in the hippocampus. Nat. Neurosci. 23(7):881-891. PubMed

Zhao X, Hsu CL, Spruston N (2022) Rapid synaptic plasticity contributes to a learned conjuctive code of position and choice-related information in the hippocampus. Neuron. 110(1):96-108.e4. PubMed

Invariant stimulus decoding using correlated neuronal fluctuations. Ebrahimi et al., (2022) Nature. May 605(7911):713-721. PubMed.

Current studies in neuroscience are carried out by examining neuronal firing as a unitary property of the nervous system. Patterns of neuronal firing called neural population codes are used to make correlations with different brain functions. It is also used to make representations of sensory perception using animal behavior in response to sensory stimuli. By analyzing neuronal firing in response to a specific stimulus over different timescales, it is found that variations of elements (neurons that fire) occur within each set of neurons that fire (Rumyantsev et al., 2020; Driscoll et al., 2017; Montijn et al., 2016). A recent study (Ebrahimi et al., 2022) shows occurrence of a) sensory coding redundancy near the beginning of perception of a sensory stimulus, and b) shared co-fluctuations of neuronal firing in different areas of brain.

To decode behavioral response and to progress from representation to causation, it is necessary to understand a mechanistic explanation how first-person property of sensory perception is generated and how it is associated with firing of different sets of neurons. Since sensory perception occurs in a conscious mind, it is necessary to examine how first-person properties occur within the nervous system and how this mechanism is correlated with third person observation such as firing of neurons. Studies have shown that oscillating extracellular potentials need to be maintained in a narrow range for conscious perception. Oscillating extracellular potentials is a reflection of ionic changes occurring across neuronal cellular membranes that in turn reflect the nature of propagation of potentials across the neuronal processes. Since oscillations across three-dimensional space of extracellular matrix (ECM) can only be explained by the occurrence of vector components that contribute to these oscillations, it is necessary to find mechanisms that lead to generation of these vector components. Since there are oscillations of potentials with different amplitudes and frequencies in space, it is also necessary to explain how and where the vector components contributing to these oscillations occur. This also provides an opportunity to hypothesize mechanism/s that can lead towards a solution for the system that can explain how first-person properties are formed within the system.

Since inner sensations of memories are first-person properties, it is possible to ask, “What type of a change should occur within the system during associative learning that can be used to generate first-person inner sensations of retrieved memories?” Once it becomes possible to generate a hypothesis for such a mechanism, it allows us to test for the occurrence of a change during learning. With this aim, semblance hypothesis synthesized a general framework of a mechanism (Vadakkan, 2007). When attempts are made to generate artificial intelligence by transferring the mechanism of natural intelligence to engineered systems, it becomes necessary to understand how first-person properties are generated within the system. Towards this attempt, memories were viewed as hallucinations (inner sensation of a sensory stimulus in its absence) and a framework for a mechanism was developed (Minsky, 1980). When semblance hypothesis was further examined in line with K-lines proposed by Marvin Minsky, it was possible to derive formation of inter-postsynaptic functional LINK (IPL) as linchpin change occurring during learning whose reactivation is expected to generate internal sensation of memory (Vadakkan, 2013). Accordingly, reactivation of IPLs from a lateral direction by a specific cue stimulus is capable of generating units of inner sensations. Propagation of potentials through established IPLs provides one of the vector components to oscillating extracellular potentials at the locations where postsynaptic terminals of the same neuronal order interact with each other in their orthogonal organization with respect to linear orientation of neurons in the consecutive neuronal orders. Synaptic transmission between linearly-oriented neurons of different neuronal orders provide the second vector component perpendicular to that occur through the IPLs.

Many neurons are held at subthreshold activation states. Background subthreshold activation of a neuron depends on natural environmental stimuli (for example, gravity), phase of oscillation of oscillating extracellular potentials and spatial and temporal summations of potentials. Using propagation of potentials that contribute vector components, it is possible to explain both generation of oscillating extracellular potentials and addition of potentials to several neurons that are held at subthreshold activation levels. In other words, stimulus under investigation provides additional EPSPs to different sets of neurons that are being held at subthreshold activation states and allows them to fire action potentials. This is demonstrated in Fig.1.

Figure.1. Sensory coding redundancy explained using an example. When a stimulus arrives, it will provide sufficient stimulus to several first order neurons that leads to their firing. Action potential triggered by an excitatory neuron will lead to synaptic transmission at the synapses on all its axonal terminals. The EPSP generated at a postsynaptic terminal gets spatially or temporally summated with the rest of the ESPSs arriving at the axonal hillock. Depending on whether the net summated EPSP crosses the threshold for firing, the postsynaptic neuron either fires or does not fire. This is pictorially depicted by the example of three neurons A, B and C that receive sub-threshold activations short of two EPSPs at their baseline resting states. A specific stimulus under investigation is marked “S1”. It provides EPSPS to all three neurons A, B and C. EPSPs 2, 3 and 4 reaching neurons A, B and C respectively arrive from either internal or external stimulus at the same time. If neuron A receives EPSP 1 and 2 simultaneously, it will lead to its firing. If neuron B receives EPSP 1 and 3 simultaneously, it will lead to its firing. If neuron C receives EPSP 1 and 4 simultaneously, it will lead to its firing. Hence, when EPSP1 arrives, firing of neurons A, B and C depends on whether they are receiving additional EPSPs concurrently or temporally so that these neurons fire. If D is a neuron of the second order of neurons and if it is being held at subthreshold state short of one EPSP and if it has inputs from neurons A, B, and C, then firing of either one of the neurons A or B or C will cause its firing. Hence, a stimulus under examination can cause firing of sets of A and D or A, B and D or A, B, C and D or B and D or B, C, and D or C and D, or A, C and D simultaneously.

Second explanation is needed for the observation of sensory coding redundancy at the start of perception (Ebrahimi et al., 2022). Redundancy of inputs is expected to minimize the effect of variations in the sets of neurons that fire. But the stochastic nature informs that something new is taking place within the circuitry. This provides a unique opportunity to examine any proposed hypothesis of brain functions for its explanatory capabilities. Since any set of nearly 140 input signals arriving through nearly tens of thousands of input terminals of a pyramidal neuron in the cortex can fire a neuron (Palmer et al. 2014; Eyal et al., 2018), there is presence of extreme degeneracy of input signals in firing a neuron (Vadakkan, 2019). Since many neurons are being held at subthreshold activation levels in the background state, and since there is presence of continuously varying internal stimuli originating from within the system, arrival of different combinations of input signals can lead to firing of the same neuron. By extension, it I possible to infer that stimulus from a sensory stimulus under examination can generate potentials that can reach neurons where they get summated with potentials from a) input signals generated either internally or externally, and b) reactivation of IPLs that also contribute to oscillating extracellular potentials. As the interval between testing increases, occurrence of different associative learning events will add more IPLs to the system. In addition, some IPLs will get reversed back over time. These can lead to changes in the net EPSPs arriving at the axonal hillocks of neurons that are held at subthreshold activation states. This can explain variations of neuronal sets that fire in response to a specific stimulus over different timescales (Rumyantsev et al., 2020; Driscoll et al., 2017; Montijn et al., 2016).

Thirdly, it is necessary to explain how different areas of brain show shared co-fluctuations of neuronal firing. As explained in the previous paragraph, addition and deletion of IPLs over time will lead to changes in the sets of neurons that fire. Both propagation of potentials along projection neurons between different brain areas, and maintenance of both frequency and amplitude of waveforms of oscillating extracellular potentials are expected to allow maintenance of correlated subthreshold states of sets of neurons at two locations. When a stimulus arrives, this can provide inputs to subthreshold-activated neurons at those two locations and allow them to cross thresholds for firing.

Both correlated fluctuations and visual coding redundancy that are time-varying throughout stimulus presentation rise within 100ms and peak around 200ms after sensory stimulus onset (Ebrahimi et al., 2022). This time delay matches with the time needed for expansion of several spines that eventually leads to the formation of more IPLs. These late-forming IPLs have no role in early perception. However, their physiological utility is in maintaining continuity of perception of a stimulus. The inference made in the work that some neurons have greater intrinsic variability in the fidelity of stimulus encoding than others can be explained by 1) Different combinations of inputs add potentials that will allow the summated EPSPs to cross the threshold to fire a neuron, and 2) IPLs are formed in excess such that same semblance can be generated from different combinations of units of semblance generated at those IPLs. Inference from all the observations tempted the authors to speculate for presence of “non-interfering communication channels” in the neocortex, which can be explained in terms of the IPL mechanism.

Driscoll LN, Pettit, NL, Minderer M, Chettih SN, Harvey CD (2017) Dynamic reorganization of neuronal activity patterns in parietal cortex. Cell 170:986-999. PubMed

Ebrahimi S, Lecoq J, Rumyantsev O, Tasci T, Zhang Y, Irimia C, Li J, Ganguli S, Schnitzer MJ (2022) Emergent reliability in sensory cortical coding and inter-area communication. Nature 605(7911):713-721. PubMed

Eyal G, Verhoog MB, Testa-Silva G, Deitcher Y, Benavides-Piccione R, DeFelipe J, de Kock CPJ, Mansvelder HD, Segev I (2018) Human cortical pyramidal neurons: From spines to spikes via models. Front. Cell Neurosci. 12:181. PubMed

Montijn JS, Meijer GT, Lansink CS, Pennartz CM (2016) Population-level neural codes are robust to single-neuron variability from a multidimensional coding perspective. Cell Rep. 16:2486-2498. PubMed

Palmer LM, Shai AS, Reeve JE, Anderson HL, Paulsen O, Larkum ME (2014) NMDA spikes enhance action potential generation during sensory input. Nat. Neurosci. 17(3):383-390 PubMed

Rumyantsev OI, Lecoq JA, Hernandez O, Zhang Y, Savall J, Chrapkiewicz R, Li J, Zeng H, Ganguli S, Schnitzer MJ (2020) Fundamental bounds on the fidelity of sensory cortical coding. Nature 580:100-105. PubMed

Mosquito brains encode unique features of human odour to drive host seeking. Zhao et al., (2022) Nature. May 605(7911):706-712. Article

Human odor stimulus leads to activation of a specific glomerulus in Aedes mosquitos. Every glomerulus receives more than on sensory neuronal input. For example, in Drosophila melanogaster a single glomerulus that senses CO₂ has more than one sensory neuron arriving to that glomerulus (Jones et al., 2007). Close examination of the findings in a recent work (Zhao et al., 2022) shows that more than one sensory neuron is necessary for a specific sensory perception to occur. It is known that neurons that express the same complement of ligand-specific receptors send axons to a single olfactory glomerulus (Vosshall & Stocker, 2007). Hence, it is generally thought that glomerulus is an ideal location to study sensory perception (Wang et al., 2003; Semmelhack & Wang, 2009).

Since more than one sensory neuron (olfactory neuron) is needed for perception to occur, it is reasonable to assume the presence of an interactive change occurring between these neurons or their immediate output neurons. This matches with the previous explanation for the generation of first-person property of perception by the semblance hypothesis (Fig.1; Vadakkan, 2015).

Based on the semblance hypothesis, units of internal sensation of perception namely perceptons are generated at the locations where two sensory inputs converge (Vadakkan, 2015). But the question is how does the fly recognize human odor for the first time as something beneficial? Generation of the first-person property of internal sensation concurrent with motor actions to fly towards humans during first instance is most likely to occur automatically by virtue of an inherited wiring mechanism. Hence, the first instance of flight towards a human most likely occurs as expected from an automaton. This and future events of flights towards humans can lead to associations between sensory inputs taste of the blood or filling of stomach or satiety can lead to formation of IPLs that can generate both internal sensations necessary for survival concurrent with appropriate motor actions. Hence, during later times, this becomes a learned behavior.

Figure 1. Schematic diagram showing the mechanism of olfactory percept formation within a glomerulus. a) Spread of activity through the neuronal processes in the absence of odorants. The baseline firing of the olfactory receptor neurons (ORNs) leads to spread of activity to the synapses between the ORNs and the projection neurons (PNs). Spread of activity through the excitatory local neurons (ELNs) from one glomerulus to other glomeruli results in oscillating activity across different glomeruli in the antennal lobe. Two postsynaptic terminals each from the corresponding three different sister PNs whose dendrites are located within a single glomerulus are shown. Based on the present work, existing inter-postsynaptic LINKs within each of the different glomeruli can contribute to horizontal component that can trigger oscillations of potentials among the glomeruli. The integral of all the non-specific semblances induced at the inter-postsynaptic LINKs is called C-semblance that can contribute to the attention of the fly. A and C are the presynaptic terminals of the ORNs. B and D are the postsynaptic terminals (dendritic spines) of two different PNs within a glomerulus. b) Induction of perceptons in the presence of an odorant. Two synapses between two ORNs and two sister PNs within the glomerulus along with their interpostsynaptic LINK (IPL) B–D is shown. In the context of background C-semblance, the stimulus-semblion U-loops form at the inter-postsynaptic LINK B–D to induce perceptons. Note that the semblions are shown to overlap closer to the olfactory receptors than the actual source of the odorant. This enables localization of the odor close to the olfactory receptors, in contrast to the visual perception. The entanglement of perceptons provides the conformation for the percept of a specific smell. Percept of a specific attractive smell formed within a glomerulus can trigger motor actions to the fly along the concentration gradient as a response to increasing percepts, the fly can reach towards the source of food. Note that the oscillating potential wave form that extend beyond the single glomerulus in the absence of odorants gets limited to that glomerulus alone due to the spread of inhibitory activity to the other glomeruli through the inhibitory local neurons (ILNs) during perception (Figure from Vadakkan, 2015).

Jones WD, Cayirlioglu P, Kadow IG, Vosshall LB (2007) Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature 445(7123):86-90. PubMed

Semmelhack JL & Wang JW (2009) Select Drosophila glomeruli mediate innate olfactory attraction and aversion. Nature 459:218-223. PubMed

Vadakkan KI (2015) A framework for the first-person internal sensation of visual perception in mammals and a comparable circuitry for olfactory perception in Drosophila. Springerplus 4:833. PubMed

Vosshall LB. & Stocker RF (2007) Molecular architecture of smell and taste in. Drosophila. Annu. Rev. Neurosci. 30:505-533. PubMed

Wang JW, Wong AM, Flores J, Vosshall LB, Axel R (2003) Two-photon calcium imaging reveals an odor-evoked map of activity in the fly brain. Cell 112:271-282. PubMed

Zhao Z, Zung JL, Hinze A, Kriete AL, Iqbal A, Younger MA, Matthews BJ, Merhof D, Thiberge S, Ignell R, Strauch M, McBride CS (2022) Mosquito brains encode unique features of human odour to drive host seeking. Nature. 605(7911):706-712. PubMed

Synaptic correlates of associative fear memory in the lateral amygdala. Choi et al., (2021) Neuron. Sept. 109(17):2717-2726. Article

Experiments have shown that activated ensembles of synapses have significantly larger spine morphology in the auditory cortex-to-lateral amygdala synaptic region after auditory fear conditioning. Fear extinction reversed these ensembles of enlarged spines, whereas re-conditioning with the same tone and shock restored the spine size of the synapses (Choi et al., 2021). In fear conditioning experiments, foot shock is a high energy stimulus, and it is likely to generate several inter-postsynaptic functional LINKs (IPLs) in the amygdala in a time-dependent manner. This can explain why fear conditioning is observed after three hours following foot shock (Rumpel, 2005). Furthermore, it was observed that memory is reduced after three hours of blocking of synaptic incorporation of AMPA receptors in as few as 10 to 20% of lateral amygdala neurons (Rumpel, 2005). This needs a timescale-matched explanation. Based on the semblance hypothesis, spine enlargement takes place at the time of associative learning. Enlarged spines can promote IPL formation that takes place in milliseconds. Intra-spine GluR1 vesicle fusion to the lateral spine head membrane that incorporates membrane segments to the lateral spine head region and enables IPLs to advance to more stabilizable states. If this is blocked, then it will lead to reversal of formed IPLs at specific locations of convergence of signals from associatively learned stimuli. It also matches with the delay in the induction of LTP after stimulation (Vadakkan, 2019). Another inference is that internal sensation of memory results from net semblance induced at a minimum number of inter-LINKed spines.

Choi DI, Kim J, Lee H, Kim JI, Sung Y, Choi JE, Venkat SJ, Park P, Jung H, Kaang BK (2021) Synaptic correlates of associative fear memory in the lateral amygdala. Neuron. S0896-6273(21)00502-X. PubMed

Rumpel S, LeDoux J, Zador A, Malinow R (2005) Postsynaptic receptor trafficking underlying a form of associative learning. Science. 308(5718):83-8. PubMed

Vadakkan KI (2009) A potential mechanism for first-person internal sensation of memory provides evidence for the relationship between learning and LTP induction. Behav Brain Res. 360:16-35. PubMed

Selective filtering of excitatory inputs to nucleus accumbens by dopamine and serotonin. Christoffel et al., (2021 PNAS.118(24):e2106648118. PubMed.

Dopamine reduces excitatory postsynaptic currents (EPSCs) generated by paraventricular thalamus (PVT) inputs to NAc, when carried out by whole cell recording from medium spiny neurons (MSNs) of NAc (Christoffel et al., 2021). This naturally leads to the question, “What mechanistic explanation can satisfy the inference that dopamine filter excitatory inputs to NAc?” Based on IPL mechanism, formation of IPLs between dendritic spines of MSNs that synapse with excitatory inputs from PVT neurons and dendritic spines of MSNs that synapse with inhibitory inputs from ventral tegmental area (VTA) takes place when dopaminergic inputs from VTA cause expansion of spines of MSNs that synapse with excitatory inputs (Vadakkan, 2019). The net effect will provide results equivalent to filtering of excitatory inputs to NAc by dopamine.

Christoffel DJ, Walsh JJ, Hoerbelt P, Heifets BD, Llorach P, Lopez RC, Ramakrishnan C, Deisseroth K, Malenka RC (2021) Selective filtering of excitatory inputs to nucleus accumbens by dopamine and serotonin. Proc Natl Acad Sci U S A.118(24):e2106648118. PubMed

Vadakkan K.I (2019) Internal sensation of pleasure can be explained as a specific conformation of semblance: Inference from electrophysiological findings. Peerj Preprints Article

Drift in the set of neurons in the primary olfactory cortex that fire in response to an odour. Schoonover et al., (2021) Nature. 594(7864):541-546. Article

Several studies have observed correlation between odorants and specific sets of neurons that fire in response to them. Continuous recording from these neurons show that this correlation is lost after several weeks (Schoonover et al., 2021; Marks and Goard, 2021; Deitch et al., 2021). Authors suspected that this instability reflects the unstructured connectivity of piriform cortex. What property of the circuitry will cause such a drift? It further leads to more fundamental questions such as “What is a percept?” “Where is it formed?” It was possible to explain a framework of a mechanism of perception based on the IPL mechanism (Vadakkan, 2011). During associative learning events, new IPL are formed in the cortices. Even though olfactory stimuli propagate directly to the hippocampus without propagating to an intermediate association cortex (Zhou et al., 2021), outputs from the hippocampus can generate IPLs in the cortex. Insertion of new neurons in the pathways (in the granule layer of hippocampus) through which signals from associatively learned items/events propagate, along with exposure of the system to new associative learning items/events that share elements of the previously associated items/events, will lead to continuous formation of new IPLs in the cortices (Vadakkan, 2010; 2016). This will lead to changes in the summated potentials arriving to the neurons in the olfactory cortex. Hence, firing property of neurons in the primary olfactory cortex during perception of the same stimulus is expected to show continuous drift.

Based on the semblance hypothesis, when perception is viewed as first-person internal sensations, it was possible to find a framework of a mechanism for perception (Vadakkan, 2015). Accordingly, internal sensation of a percept is formed by integral of all perceptons, unitary mechanisms of perception. Large number of redundant perceptons are expected to form. Hence, the net integral of all perceptons remain almost same, even with changes in the locations from where perceptons are formed. Furthermore, extreme degeneracy of attenuating input signals in firing a neuron (Vadakkan, 2019) indicates that perceptons are generated at the input level. Correlations with neuronal firing will only be true for those neurons that are being held at sub-threshold activation state and receive additional potentials through inter-postsynaptic functional LINKs (IPLs) at the time of perception. Hence, internal sensation of perception continues to take place even when the set of neurons that fires changes over time due to changes in the circuitry.

Deitch D, Rubin A, Ziv Y (2021) Representational drift in the mouse visual cortex. Curr. Biol. 31, 4327–4339. PubMed

Marks TD & Goard MJ (2021) Stimulus-dependent representational drift in primary visual cortex. Nat. Commun. 12, 5169. PubMed

Schoonover CE, Ohashi SN, Axel R, Fink AJP (2021) Representational drift in primary olfactory cortex. Nature. 2021 594(7864):541-546. PubMed

Vadakkan KI (2011) A possible mechanism of transfer of memories from the hippocampus to the cortex. Med Hypotheses. 77(2):234-43. PubMed

Vadakkan KI (2015) A framework for the first-person internal sensation of visual perception in mammals and a comparable circuitry for olfactory perception in Drosophila. Springerplus. 4:833. PubMed

Vadakkan KI (2016) The functional role of all postsynaptic potentials examined from a first-person frame of reference. Rev Neurosci. 27(2):159-84. PubMed

Vadakkan KI (2019) Extreme degeneracy of inputs in firing a neuron leads to loss of information when neuronal firing is examined. Peerj Preprints. Article

Zhou G, Olofsson JK, Koubeissi MZ, Menelaou G, Rosenow J, Schuele SU, Xu P, Voss JL, Lane G, Zelano C (2021) Human hippocampal connectivity is stronger in olfaction than other sensory systems. Prog Neurobiol. 201:102027. PubMed

Pathological features of Alzheimer's disease such as tangles & plaques start appearing in normally aging brains (Was able to explain this recently: Aging as a loss of an adaptation that stabilizes last developmental stage of the nervous system. 

Examination of IPLs derived by semblance hypothesis has led to the inference that the last stage of its development undergoes an adaptation whereby inter-neuronal inter-spine fusion is prevented by arresting it at/before the stage of hemifusion (Vadakkan, 2020). This is based on the following observations. In the mouse, neuronal precursor cells in the ventricular zone (VZ) undergo cell division. While in the VZ, 100% of precursors in G2 and S phases of the cell cycle couple together and form clusters (Bittman et al., 1997). During this stage, injection of dye into one cell spread to neighouring cells (Bittman et al., 1997). This indicates the formation of fusion pores between these cells. This is followed by death of nearly 70% of these cells and survival of the remaining 30% cells (Blaschke et al., 1996). The surviving 30% of cells are expected to have acquired an adaptation, most probably during inter-cellular coupling. The adaptation most likely prevents any future coupling between neurons that may result in inter-neuronal fusion. This adaptation is suitable for maintaining IPLs (that generated inner sensations) and prevents any IPL fusion. Aging can be viewed as resulting from gradual loss of this adaptation. Augmented formation IPL fusion events can lead to pathological changes such as those observed in neurodegenerative disorders (Vadakkan, 2019). For example, pathological changes of neurofibrillary tangles and amyloid plaques can result from precipitation of proteins & leakages of certain precipitated proteins through defective fusion pores to the extracellular matrix space in Alzheimer’s disease. If the semblance hypothesis is correct, then its corollary that these pathological findings should also be found in normal aging can be verified. Since senile neurofibrillary tangles and amyloid plaques appear in normally aging brains (Anderson, 1997; Saha and Sen, 2019), this forms sufficient verification. This reinforces the need for testing the predictions of semblance hypothesis.

Anderton BH (1997) Changes in the ageing brain in health and disease. Philos Trans R Soc Lond B Biol Sci. 352(1363):1781-1792. PubMed

Bittman K, Owens DF, Kriegstein AR, LoTurco JJ (1997) Cell coupling and uncoupling in the ventricular zone of developing neocortex. Journal of Neuroscience 17(18):7037-7044. PubMed

Blaschke AJ, Staley K, Chun J (1996) Widespread programmed cell death in proliferative and postmitotic regions of the fetal cerebral cortex. Development 122(4):1165-74. PubMed

Saha P, Sen N (2019) Tauopathy: A common mechanism for neurodegeneration and brain aging. Mech Ageing Dev. 178:72-79. PubMed

Vadakkan KI (2020) A derived mechanism of nervous system functions explains aging-related neurodegeneration as a gradual loss of an evolutionary adaptation. Curr Aging Sci 13(2):136–152. PubMed

Vadakkan KI (2016) Neurodegenerative disorders share common features of "loss of function" states of a proposed mechanism of nervous system functions. Biomed Pharmacother. 83:412-430. PubMed

Heterogeneity of neurons in the cortex

Studies of cortical neurons show significant heterogeneity in transcriptomic analyses (Tasic et al., 2016; Cembrowski et al., 2016; Tasic et al., 2018; Hodge et al., 2019). In fact, these findings show that there won't be two neurons with the same sets of transcripts within them. The above findings naturally raise the question, "What is the functional importance of such a finding?" The actual operational mechanism of the nervous system is expected to provide clues for a suitable explanation. Based on the IPL mechanism, this heterogeneity is necessary for the formation of IPL fusion between spines that belong to different neurons at one stage of development supported by the diffusion of dye injected into on neuron to neighboring neurons (see, Vadakkan, 2020). If neurons are not heterogeneous, then fusion between them will not evoke cellular reactions, which is responsible for cell death of majority of neurons. Most importantly, this IPL fusion is expected to trigger an adaptation in surviving neurons, responsible for restricting IPL fusion to the stage of IPL hemifusion. Thus, neuronal heterogeneity can be viewed as a marker of an adaptation that occurred during those last stages of the developmental of the nervous system. It is most likely that maintaining heterogeneity is essential for maintaining the above adaptation throughout the life-span of the neurons. This prompts us to make a testable prediction that, any deficiencies in maintaining this adaptation will trigger IPL fusion between heterogeneous neurons, which can explain aging and other disease associated neurodegeneration.

Cembrowski MS, Spruston N (2019) Heterogeneity within classical cell types is the rule: lessons from hippocampal pyramidal neurons. Nat Rev Neurosci. 20(4):193-204. PubMed

Cembrowski MS, et al., (2016) Spatial gene-expression gradients underlie prominent heterogeneity of CA1 pyramidal neurons. Neuron. 89(2):351-68. PubMed

Hodge et al., (2019) Conserved cell types with divergent features in human versus mouse cortex. Nature 573 (7772):61-68. PubMed

Tasic et al., (2016) Adult mouse cortical cell taxonomy revealed by single cell transcriptomics. Nat Neurosci. 19(2):335-346. PubMed

Tasic et al., (2018) Shared and distinct transcriptomic cell types across neocortical areas. Nature 2018 563 (7729):72-78. PubMed

Vadakkan KI. (2020) A derived mechanism of the nervous system functions explains aging-related neurodegeneration as a gradual loss of an evolutionary adaptation. Curr Aging Sci. doi: 10.2174/1874609813666191223092451. PubMed

   Spine depolarization without dendritic depolarization

   It was found that in excitatory synapses, large spine depolarization recruit voltage-dependent channels without dendritic depolarization, due to high spine neck resistance (Beaulieu-Laroche and Harnett, 2018). Hence, it leads to the following questions. "What is the functional importance of seemingly isolated spine depolarization?" and "Since this is a conserved property, how to provide a mechanistic explanation in terms of brain functions?" Another finding from the same laboratory is that distal human dendrites provide limited excitation to the soma even in the presence of dendritic spikes (Beaulieu-Laroche et al., 2018). The observation that even dendritic spikes have only a limited role in neuronal firing is of huge significance. This again reinforces the need for figuring out the functions achieved by depolarization of spine heads in excitatory cortical neurons. The IPL mechanism can explain how depolarization of spines is associated with generation of units of internal sensations independent of neuronal firing. These experimental findings compel us to undertake dedicated experimental verification of the IPL mechanism.

Beaulieu-Laroche L and Harnett MT. 2018. Dendritic spines prevent synaptic voltage clamp. Neuron 97(1): 75–82.e3. PubMed

Beaulieu-Laroche L, Toloza EHS, van der Goes MS, Lafourcade M, Barnagian D, Williams ZM, Eskandar EN, Frosch MP, Cash SS, Harnett MT. 2018. Enhanced dendritic compartmentalization in human cortical neurons. Cell 175(3): 643–651.e14. PubMed

Largest class of neurons in the visual cortex is not reliably responsive to any of the visual stimuli

In a recent report by de Vries et al., (de Vries et al., 2020), the authors examined firing of nearly 60,000 visual cortical neurons in response to different visual stimuli. They found that while most classes of these neurons respond to specific subsets of stimuli, the largest class is not reliably responsive to any of the stimuli. The latter finding supports the observations made by semblance hypothesis during visual perception (Vadakkan, 2016). Accordingly, the internal sensation of perception takes place at the inter-LINKed spines and is independent of firing of their neurons. Moreover, postsynaptic potentials generated by visual stimuli at these inter-LINKed spines need not necessarily add potentials to raise the summated potentials to reach the threshold level for firing those neurons (Vadakkan, 2019). Therefore, as per semblance hypothesis, the expectation is that a huge set of neurons will not be responsive to any visual sensory stimuli even when internal sensation of vision takes place. Report by de Vries et al., (de Vries et al., 2020) is in agreement with the expectations of the mechanism of visual perception provided by semblance hypothesis.

Their finding that most classes of visual cortical neurons respond to specific subsets of stimuli indicates that the propagation of stimuli to higher cortical areas is necessary for performing secondary functions such as a) “where” and “what” associative properties of visual stimuli at higher cortical areas, and b) associative learning with other sensory stimuli at different associative cortical areas. Due to extreme degeneracy of inputs in firing a cortical neuron (Vadakkan, 2016), two findings are expected. a) a specific neuron will respond to a very large number of visual stimuli if that neuron is being kept at sub-threshold activation level at the baseline state, and b) internal sensation of perception will continue to occur at the inter-LINKed spine of a neuron even without any change in the firing status of that neuron which remains at a supra-threshold activation state.

de Vries et al., (2020) A large-scale standardized physiological survey reveals functional organization of the mouse visual cortex. Nat Neurosci. 2020 Jan;23(1):138-151. doi: 10.1038/s41593-019-0550-9. PubMed

Vadakkan KI (2016) A framework for the first-person internal sensation of visual perception in mammals and a comparable circuitry for olfactory perception in Drosophila. Springerplus. 2015 Dec 30;4:833. doi: 10.1186/s40064-015-1568-4. eCollection 2015. PubMed

Vadakkan KI (2019) Extreme degeneracy of inputs in firing a neuron leads to loss of information when neuronal firing is examined. Peerj Preprints. Article

Artificial firing of a neuron leads to firing of a set of neurons of the same neuronal order

In a recent work by Chettih and Harvey (Chettih & Harvey, 2019), authors artificially triggered several spikes (action potentials) in single neurons in layer 2/3 of mouse visual cortex V1area. This resulted in spiking activity in a group of sparsely distributed neighboring neurons in the same neuronal order and were correlated in time. The small population of neurons that were excited were located at short distance (25–70µm) from the stimulated neuron. The stimulation had no influence beyond 300µm (for a summary, see News and Views article by Ikuko Smith (Smith, 2019). The authors called this lateral spread of activity between neurons "influence-mapping."

There is one important question. How does excitation reach at the laterally located neurons in a time-correlated manner, which is responsible for influence-mapping? This can be explained by the testable mechanism derived by semblance hypothesis (Fig.1). It is related to the previous explanation of visual perception as a first-person property using the derived mechanism of generation of internal sensation at physiological timescales (Vadakkan, 2016). The units of internal sensation of perception are induced at the inter-LINKed spines that belong to different neurons. When a single neuron is artificially fired, the back propagating action potentials will reach the dendritic spines. It will then continue to propagate through the inter-LINKed spines to the neuronal soma of the inter-LINKed spine’s neuron (Fig. 2). The spines that inter-LINK can belong to neurons that are separated by up to 300µm, a distance beyond which the probability of overlapping of dendritic arbor between neurons diminishes substantially.

Figure 1. Schematic diagram showing the route of propagation of action potential from the artificially fired neuron N1 towards the sparsely located neuron N2 within the layer2/3 in visual cortex. This spread taking place through the inter-LINKed spines Post1 and Post2 can explain what the authors describe as “influence-mapping.” Note that the inter-postsynaptic functional LINK (IPL) between Post1 and Post2 was explained as responsible of induction of internal sensation for perception (Vadakkan, 2015). Overlapping of the dendritic arbors between the neurons N1 and N2 increases the probability of IPL formation when neurons N1 and N2 are separated only by a short distance (25–70µm).

b) When a neuron was fired, the majority of neurons that were tuned to respond to similar features to that neuron were strongly suppressed than the neurons with a different tuning regardless of the distance from the stimulated neuron. Inhibition of the spikes in the neighboring neurons can be explained by activation of surrounding inhibitory interneurons. Burst of action potentials in excitatory neurons can activate somatostatin expressing inhibitory interneurons (Kwan and Dan 2012). Similar types of inhibition of surrounding areas are seen in locations where the internal sensation of perception is expected to occur in the olfactory glomeruli in Drosophila. When one glomerulus is activated, inhibitory local interneurons (ILN) inhibit all the remaining glomeruli (Hong and Wilson, 2015) enabling the specificity of the percept for that particular smell (Vadakkan, 2015).

Orientation tuning is tested by a source of light. This will cause activation of a large number of islets of inter-LINKed spines within one cortical column. But when single neurons are artificially fired the backpropagation of potentials will reach only specific sets of inter-LINKed spines. This explains why only neurons that are located sparsely are fired, correlated in time.

Verification: Based on semblance hypothesis, the prediction that can be made is the presence of inter-postsynaptic functional LINKs (IPLs) between spines that belong to the artificially fired neuron and the sparsely located neurons that were fired in a time-correlated manner.

Chettih SN, Harvey CD (2019) Single-neuron perturbations reveal feature-specific competition in V1. Nature doi: 10.1038/s41586-019-0997-6. PubMed

Hong EJ, Wilson RI (2015) Simultaneous encoding of odors by channels with diverse sensitivity to inhibition. Neuron 85(3):573–589. PubMed

Kwan AC, Dan Y (2012) Dissection of cortical microcircuits by single-neuron stimulation in vivo. Current Biology 22, 1459–1467. PubMed

Smith IT (2019) The influence of a single neuron on its network. Nature. 567(7748):320-321 PubMed

Vadakkan KI (2015) A framework for the first-person internal sensation of visual perception in mammals and a comparable circuitry for olfactory perception in Drosophila. Springerplus 4:833. PubMed

Memory retrieval occurs at a frequency of oscillating extracellular potentials similar to that was present during learning

A recent study examined the nature of oscillating extracellular potential both during learning and memory retrieval (Vaz et al.. 2019).

In order to reactivate the same set of IPLs that formed during learning at the time of memory retrieval, it is necessary to have almost similar conditions that were present at the time of learning. Maintaining the same frequency of oscillating extracellular potentials is a major factor in achieving this. Based on the semblance hypothesis, the synaptic transmission in one direction and propagation of potentials in a near-perpendicular direction through the inter-postsynaptic functional LINK (IPL) contribute vector components to the oscillating extracellular potentials, which is essential for binding and integration of units of internal sensations for providing the sensory qualia of memory. The findings of this study show that similar frequency of oscillating extracellular potentials are present both during learning and memory retrieval support the expectations of semblance hypothesis.

Vaz AP, Inati SK, Brunel N, Zaghloul KA (2019) Coupled ripple oscillations between the medial temporal lobe and neocortex retrieve human memory. Science. 363:975-978. PubMed

Dendritic calcium spikes that are related to behavior and cognitive function

Similar to the action potentials (axonal spikes or neuronal firing) occurring at the axonal hillock, there are spikes occurring at the dendrites. These are called dendritic spikes. Based on the strength of summated potentials, a rough estimate shows that they constitute synchronous activation of nearly 10 to 50 neighboring glutamatergic synapses triggering a local regenerative potential (Antic et al., 2010). Depending on the channels involved, there are different types of dendritic spikes. Recently, it was found that distal dendrites generate dendritic spikes whose firing rate is nearly five times greater than at the cell body (Moore et al., 2017). Another group of investigators who have previously shown that dendritic spikes are related to behavior and cognitive function recently found that dendritic calcium spikes contribute to surface potentials that are recorded as electroencephalogram (EEG) (Suzuki et al., 2017). Surface EEG recording is generated by current sink that reflects the net potential changes within the extracellular matrix space. This is expected to be contributed by several factors. It is known that the surface positive potentials are generated mainly by synaptic inputs from other cortical and subcortical regions to the pyramidal neurons located between L2/3 to L4 regions (Douglas and Martin, 2004). Recent studies by Suzuki et al., has found that dendritic calcium spikes at the main bifurcation points of the apical dendrites of L5 pyramidal neurons (note that L5 pyramidal neurons are upper motor neurons that direct motor movement of the body) also generate the surface positive potentials (Suzuki et al., 2017).

The last two findings lead to these questions. “How can two different sources of potentials provide similar surface positive potentials?" "Can we provide an interconnected explanation?" Since dendritic spikes are related to both behavior and cognitive functions and since IPL mechanism can explain generation of concurrent internal sensation of memory and behavioral motor action, can IPL mechanism explain the above findings? Since the apical tuft regions of all the pyramidal neurons are anchored to the pial surface, the dendritic arbor of all the pyramidal neurons overlapped at the recording location of Suzuki et al., (Suzuki et al., 2017). In this context, it is necessary to examine the potential changes occurring at the neuronal processes around the recording electrode. In the context of the IPL mechanism, it is anticipated that the dendritic spines of different neurons have formed a large number of islets of IPLs between them at these locations. By examining the zone from where low-threshold calcium spikes were recorded (Suzuki et al., 2017; Larkum and Zhu, 2002), the following is possible.

Spatially, the main bifurcation points of the apical dendrites of L5 pyramidal neurons are also locations where spines of the L2/3 pyramidal neurons receive their input. Based on the IPL mechanism, several of these spines are expected to be inter-LINKed to form large islets. These islets are also expected to be inter-LINKed with spines of L5 pyramidal neurons for initiating or controlling motor actions. The potentials through the IPLs are expected to arrive at the axon hillock of the L5 motor neurons that are kept at a sub-threshold state (see figure 5 in the FAQ section of this website) for the motor action (Fig.2). For a system that operates to generate internal sensations and initiates or controls concurrent motor actions, the islets at appropriate locations are expected to transmit potentials to the axon hillock of the L5 pyramidal neurons that are upper motor neurons. Calcium spikes are generated at the postsynaptic locations within the islet of inter-LINKed spines, possibly due to an increased density of these channels at these locations. Since the pyramidal neurons are found to be under the influence of an inhibitory blanket (Karnani et al., 2014), a function of dendritic spikes is to generate sufficient potentials to overcome this inhibition. In other words, there is a provision for increasing the inhibitory blanket around an L5 pyramidal neuron axon hillock as the size of the islets of inter-LINKed spines that are connected to these neurons increases. This will make sure that the L5 neuron fires only at the activation of specific sets of IPLs that generates a specific conformation of semblance for both the internal sensation and concurrent behavioral motor actions.

Figure 2. Figure explaining a potential mechanism occurring at the level of the main bifurcation point of an apical dendrite of an L5 pyramidal neuron (based on semblance hypothesis). The circles with different colors represent an islet of inter-LINKed spines (dendritic spines or postsynaptic terminals) that belong to different pyramidal neurons at the level of the main bifurcation point of the apical dendrite of L5 neuron. Note that one of the spines (in violet) belongs to one of the L2/3 pyramidal neurons. Also note that the inter-LINKed spine on the far-right end of the islet (in green) belongs to L5 pyramidal neuron. During development, neurons of different cortical neuronal orders descend from the inner pial surface area by anchoring the apical dendritic terminals to the inner pial region. This allows overlapping of the dendritic arbors of neurons from different orders, which leads to abutting of their spines that eventually leads to the formation of inter-LINKs between these spines during learning. The waveform shown at the level of the inter-LINKed spines indicates that the oscillating extracellular potentials recorded have a major contribution from the propagation of potentials through the islets of inter-LINKed spines. Secondary factors can determine different wave forms depending on the locations from where recording is carried out. They include a number of neuronal layers, recurrent collaterals, connections with the projection neurons from other areas of the brain, etc. Figure not to scale (spines in the islet are drawn disproportionately large compared to the size of neurons).

The explanation that synaptic transmission and propagation of potentials through the IPLs provide vector components of oscillating extracellular potentials also becomes suitable. If the arrival of potentials from sensory stimuli evokes dendritic calcium spikes along with the reactivation of specific inter-LINKed spines (and their islets) inducing units of specific internal sensations concurrent with activation of specific sets of motor neurons, it can provide an explanation how dendritic calcium spikes are related to behavior and cognitive function. The findings of Suzuki et al., necessitate examining the role of background EEG wave forms, frequency of which correlates with normal level of consciousness. In this regard, the explanation by the IPL mechanism that the net background semblance induced by reactivation of inter-LINKed spines contributes to the internal sensation of consciousness (Vadakkan, 2010) becomes a suitable mechanism that can be subjected to further studies.

Antic SD, Zhou WL, Moore AR, Short SM, Ikonomu KD (2010) The decade of the dendritic NMDA spike. J Neurosci Res. 88(14):2991–3001 PubMed

Douglas RJ, Martin KA (2004) Neuronal circuits of the neocortex. Annu. Rev. Neurosci. 27: 419–451 PubMed

Karnani MM, Agetsuma M, Yuste R (2014) A blanket of inhibition: functional inferences from dense inhibitory connectivity. Curr Opin Neurobiol. 26:96-102. PubMed

Larkum ME, Zhu JJ (2002) Signaling of layer 1 and whisker-evoked Ca2+ and Na+ action potentials in distal and terminal dendrites of rat neocortical pyramidal neurons in vitro and in vivo. J. Neurosci. 22, 6991–7005 PubMed

Moore JJ, Ravassard PM, Ho D, Acharya L, Kees AL, Vuong C, Mehta MR (2017) Dynamics of cortical dendritic membrane potential and spikes in freely behaving rats. Science. 355(6331) PubMed 

Suzuki M, Larkum ME (2017) Dendritic calcium spikes are clearly detectable at the cortical surface. Nat Commun. 8(1):276. PubMed

Vadakkan KI (2010) Framework of consciousness from semblance of activity at functionally LINKed postsynaptic membranes. Front Psychol. 1:168. PubMed

Regenerative spikes at the dendritic arbor - a mechanism for internal sense of a place that reflects binding at the time of learning

Each place field consists of a unique set of CA1 neurons that fire action potential. At the dendritic regions, calcium transients inform about a change in potentials occurring regeneratively either due to back propagating action potentials (bAP) or by dendritic spikes. Recent studies observed calcium transients secondary to regenerative dendritic events in place cells that can predict place field properties (Sheffield and Dombeck, 2015a; Sheffield et al., 2017). These calcium transients have a highly spatiotemporally variable prevalence throughout the dendritic arbor. In some cases, only a subset of the observed branches displayed detectable spikes, which indicates that spikes originated at these dendritic branches. None of the observed branches in many cases displayed detectable spikes during place field traversals while the soma (and axon) fired. This means that the bAP did not reach these locations. From the findings of Sheffield and Dombeck, it is clear that dendritic spikes relate to spatial precision. However, this finding needs a mechanistic explanation.

The above finding can be explained by the occurrence of dendritic spike occurs at an islet of inter-LINKed spines that belong to different CA1 neurons (Vadakkan, 2013). This has the following advantages. a) Activation of inter-LINKed spines within an islet of inter-LINKed spines induces units of internal sensations for a specific place. b) One dendritic spike at an islet of inter-LINKed spines that belong to different neurons can explain the firing of different CA1 neurons that are being maintained in a sub-threshold state at the time of the dendritic spike. It also supports why a high percentage of place cells are shared between different places. c) Since potentials degrade as they reach the axonal hillock, it may require potentials arriving from more than one spike to contribute to the firing of a CA1 neuron depending on latter’s sub-threshold level. d) The highly spatiotemporally variable nature of spike depends on the qualia of internal sensations that they induce in response to and matching with the place (which depends on previous associative learning events with different places). The latter property can explain the expected binding feature (Sheffield and Dombeck, 2015b).

Sheffield MEJ, Dombeck DA (2015a) Calcium transient prevalence across the dendritic arbour predicts place field properties. Nature. 517(7533):200-204. PubMed

Sheffield MEJ, Adoff MD, Dombeck DA (2017) Increased Prevalence of Calcium Transients across the Dendritic Arbor during Place Field Formation. Neuron. 96(2):490-504.e5 PubMed

Sheffield ME, Dombeck DA (2015b) The binding solution? Nature Neuroscience. 18(8):1060-102 PubMed

Vadakkan KI (2013) A supplementary circuit rule-set for neuronal wiring. Frontiers in Human Neuroscience. 7:170 PubMed

B. In pathological conditions

1. Spread of epileptic activity

Epileptic activity in the hippocampus propagates with or without synaptic transmission at a speed of nearly 0.1m/s (Jefferys, 2014). Experiments showed that the longitudinal propagation of epileptic activity from one end of a neuronal order to its other end in the hippocampus takes place independent of chemical or electrical synaptic transmission (Zhang et al., 2014). Since this spread of epileptic activity occurs at a speed of 0.1 m/s and is not compatible with ionic diffusion or pure axonal conduction (Jefferys 2014; Zhang et al., 2014), it requires an explanation at the cellular and electrophysiological levels. In this regard, rapid chain propagation through the inter-postsynaptic functional LINKs (IPLs) explained by the semblance hypothesis (Vadakkan, 2015) offers a suitable explanation for a mechanism.

Jefferys JG (2014) How does epileptic activity spread? Epilepsy Currents. 14(5):289-290 PubMed

Vadakkan KI (2016) Rapid chain generation of interpostsynaptic functional LINKs can trigger seizure generation: Evidence for potential interconnections from pathology to behavior. Epilepsy & Behavior. 59:28-41 PubMed

Zhang M, Ladas TP, Qiu C, Shivacharan RS, Gonzalez-Reyes LE, Durand DM (2014) Propagation of epileptiform activity can be independent of synaptic transmission, gap junctions, or diffusion and is consistent with electrical field transmission. Journal of Neuroscience. 2014 34(4):1409-1419 PubMed

2. Heterogeneity of clinical and pathological findings in Alzheimer's disease

Alzheimer's disease (and most other neurodegenerative disorders) is highly heterogeneous in its clinical and pathological features (Lam et al., 2013; Esteves and Cardoso, 2020). Since transcriptomic analysis shows that no two neurons are same (Tasic et al., 2016; Cembrowski et al., 2016; Tasic et al., 2018; Hodge et al., 2019) and since IPL formation can occur between abutted spines that belong to different neurons at locations of convergence (Vadakkan, 2019), pathological IPL fusion changes expected to occur in neurodegenerative disorders occur between different sets of neurons in different patients. Hence, depending on the outcome of damage that can occur due to the specific combinations of fusion between different sets of neurons, huge heterogeneity can be expected.

Cembrowski MS, Bachman JL, Wang L, Sugino K, Shields BC, Spruston N (2016) Spatial gene-expression gradients underlie prominent heterogeneity of CA1 pyramidal neurons. Neuron. 89(2):351-68. PubMed

Esteves AR, Cardoso SM (2020) Differential protein expression in diverse brain areas of Parkinson’s and Alzheimer’s disease patients. Sci. Rep. 2020, 10:1–22. PubMed

Hodge et al., (2019) Conserved cell types with divergent features in human versus mouse cortex. Nature 573 (7772):61-68. PubMed

Tasic et al., (2016) Adult mouse cortical cell taxonomy revealed by single cell transcriptomics. Nat Neurosci. 19(2):335-346. PubMed

Lam B, Masellis M, Freedman M, Stuss DT, Black SE. (2013) Clinical, imaging, and pathological heterogeneity of the Alzheimer's disease syndrome. Alzheimers Res Ther. 2013 Jan 9;5(1):1 PubMed

Tasic et al., (2018) Shared and distinct transcriptomic cell types across neocortical areas. Nature 2018 563 (7729):72-78. PubMed