Classical and Non-Classical Neural Communications

William Winlow ^1,2,*, Rouholah Fatemi ³, Andrew S. Johnson ¹

1. Dipartimento di Biologia, Università degli Studi di Napoli, Federico II, Via Cintia 26, 80126 Napoli, Italia

2. Institute of Ageing and Chronic Diseases, The Apex Building, West Derby Street, University of Liverpool, Liverpool, L& 8TX, UK

3. Department of Sport Physiology, Dehdasht Branch, Islamic Azad University, Dehdasht, Iran

* Correspondence: William Winlow

Academic Editor: Severn B. Churn

Special Issue: Neuroscience and Information Technology

Received: April 13, 2023 | Accepted: August 21, 2023 | Published: August 27, 2023

OBM Neurobiology 2023, Volume 7, Issue 3, doi:10.21926/obm.neurobiol.2303181

Recommended citation: Winlow W, Fatemi R, Johnson AS. Classical and Non-Classical Neural Communications. OBM Neurobiology 2023; 7(3): 181; doi:10.21926/obm.neurobiol.2303181.

© 2023 by the authors. This is an open access article distributed under the conditions of the Creative Commons by Attribution License, which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original work is correctly cited.

1. Introduction

Motor command elements exist in nervous systems with many examples in both vertebrates and invertebrates. The exact nature of those elements depends on the complexity of the nervous system in question. For example human nervous systems with 86 billion neurons [1] might be expected to be differently organized from those of invertebrates such as the pulmonate mollusk Lymnaea stagnalis (L.) with only about 25,000 neurons [2] or the lobster Homarus gammarus (L) with an estimated 100,000 neurons [3]. Motor output leading to locomotion and escape behavior of many sorts is a primary function of nervous systems and in vertebrates command systems of interneurons send signals via tracts of axons in the spinal cord to activate central pattern generators to instigate movements which may then be modulated by external events and appropriate feedback systems [4]. The three major components of motor commands are: excitation, inhibition and neuromodulation [5]. Because of the relative paucity of neurons in invertebrates, it was assumed that single command interneurons could initiate some specific behaviour patterns [6], but these were often described in tethered [7] or reduced preparations such as the crayfish abdomen [8,9,10]. However, Wine [11] demonstrated that the tail flip’s escape circuitry, although exhibiting a high degree of autonomy, had several motor systems involved, with 90 pairs of efferents and 18 pairs of interneurons participating. This implies parallel distributed processing (see below) so that no individual neuron could act as a command neuron. However, some neurons had dual behavioral functions [12], by which time they were described as “elements in a command group”, perhaps interacting with several central pattern generators (CPGs). The command neuron concept still has adherents or at least those who use the terminology [13,14,15]. However, in more complex animals, many now prefer to use terms such as command-like neurons, command systems of neurons or command areas in the nervous system [4]. However command neurons are now described in insects such as the fruit fly Drosophila melanogaster, where a single pair of interneurons command its feeding motor program [16] and a pair of commissural command neurons have been shown to induce wing grooming [17]. Furthermore, cells that may be command neurons control walking in the cricket Gryllus bimaculatus were described in 1992 [18]. Two neurons transmitting sky compass command signals to the motor center in the brain of the desert locust Schistocerca gregaria and numerous other descending brain neurons have been described in this animal [19]. Arguments about motor command are still lively.

Modern neuroscience concepts strongly suggest that individual neurons have a computational function [20,21] as well as their traditional functions of communication and modulation [22]. Furthermore, the brain may function as a quantum phase computer [23]. The first evidence for quantum entanglement in the human nervous system, possibly underlying consciousness, has recently been presented [24].

2. Methodology, Resources and Literature Searches

In preparing this article we reviewed literature acquired from the NIH National Library of Medicine – PubMed- and more generalized web searches as well as W Winlow’s substantial publication archive collected over 56 years from 1967. Most of the work reviewed here was from laboratories using standard electrophysiological intracellular and extracellular recording techniques, but later work uses cell culture and molecular biological techniques. The work on conscious human beings was carried out using nuclear magnetic resonance (NMR) detection and magnetic resonance imaging (MRI).

3. The Attractiveness of the Command Neuron Concept

The command neuron concept was first introduced by Wiersma and Ikeda [6] to describe specific interneurons, which when directly stimulated, were capable of generating motor actions similar to elements of normal behavior in crayfish and related species. It was suggested by Wiersma and Ikeda [6] that single neurons could alter the beat of swimmerets in crayfish. Further reports of command neurons came from a variety of sources, mainly in arthropods and mollusks (e.g. [8,9,10,25,26,27,28,29,30,31]) the Mauthner cells in fish and amphibians (e.g. [32]) and the squid giant fiber system [33]. Many workers adhered to this conclusion and examples of command neurons have been described or assumed in the following years (e.g. [15]). However, even in invertebrates, neural redundancy occurs, where multiple neural units can have similar or identical functions, leading to questioning the command fiber concept by Kupferman and Weiss [34]. They suggested that a command neuron must be both necessary and sufficient to induce the behavior ascribed to it. In addition, a command neuron has also been defined as an interneuron “reproducibly releases motor action resembling a major piece of normal behaviour upon stimulation. Other than a neuron early in an afferent pathway or late in an efferent pathway” [35]. Thus a command neuron should be both necessary and sufficient and be a true interneuron, but was the definition of an “interneuron” appropriate in all cases?

The earliest examples of presumed command neurons were described in the ladder-like nerve cords of arthropods where they were easiest to find. However, they only generated incomplete segments of full behavioral repertoires. More recently hermaphrodite specific command neurons (HSNs) have been described in Caenorhabditis elegans [36,37], which has only 302 neurons, all specified. The HSNs are multifunctional, both serotonergic and neuropeptidergic, and are also involved in other behaviors such as high-speed locomotion [38]. Command-like neurons may also occur in mollusks, which do not have ladder-like nervous systems, but whose torted CNS may have been derived from archetypal bilaterally symmetrical proto-mollusks [39,40,41,42,43] perhaps with ladder-like cords, as in the archaeogastropod Haliotis [39]. However, two extremes of neural development occur in this phylum when comparing gastropods and cephalopods (see also Figure 1). As mentioned above, there is a paucity of neurons in gastropods such as Lymnaea (25,000 neurons) and its relatives. In contrast, the nervous systems of cephalopods such as Octopus vulgaris and the squid, Sepioteuthis lessoniana, may contain 520 million neurons – (300 million of which are in the arms-Young 1963).

Figure 1 A representation of the increasing complexity of nervous systems. As numbers of neurons increase there is a significant increase in their redundancy, reducing the requirement for individual command neurons and multiplying the number of feedback loops in the system. Colours represent animals in same phylum: Red for chordates (vertebrates); Green for arthropods; blue for molluscs; orange for the annelid H. medicinalis and black for the free-living nematode roundworm C. elegans [37,44,45,46].

3.1 Gastropod Mollusks

Gastropod mollusks such as Helix pomatia, Lymnaea stagnalis and the related species Helisoma trivolvis and Planorbis corneus have been studied in detail because they possess neurons with giant cell bodies which are relatively easy to penetrate with glass microelectrodes. Recordings from these cells have shown their role in the control of behaviors such as feeding [47], locomotion [48,49], respiration [50,51,52], defensive behaviors [53], etc. To take a specific example, in Lymnaea stagnalis, the giant multi-action neuron right pedal dorsal, RPeD1 [54] might be considered to be a command neuron [55,56] as originally envisaged by Wiersma and Ikeda [6]. RPeD1 is a key component of the central respiratory generator of Lymnaea, which has been cultured in isolation ([57] and see Figure 2A) and which, in semi-intact preparations, can trigger the opening of the pneumostome in the absence of motor neuronal activity [55] as well as having direct synaptic connections to the motor neurons (Figure 2C). It thus acts as a motor neuron and an interneuron and participates in the central pattern generator for aerial respiration. Thus RPeD1 fulfils the sufficiency criterion for being a command neuron, but is it also necessary to fulfil its behavioral role?

Figure 2 Connectivity in the brain of Lymnaea stagnalis (L.), an animal with only about 25,000 neurons in its nervous system, although its connectome is not yet completed. A) the neurons of the respiratory central pattern generator (RCPG), arrows indicate excitatory connections, while ovals indicate inhibitory connections: RPeD1- right pedal dorsal neuron 1; VD4 – visceral dorsal neuron 4; Ip3I – input 3 interneuron. B) A diagram of the brain, except for the buccal ganglia in which much of the feeding system is located. The RCPG neurons identified by colour as in A), showing its dispersed nature across three separate ganglia (modified from Winlow et al, 2018 [58]). C) A circuit diagram of the distributed, interacting systems within the brain of Lymnaea. Arrows again indicate excitatory synapses while ovals indicate inhibitory connections. Dotted lines indicate presumed connections. (From Winlow and Polese 2014 [43], with permission from Nova Science Publishers, Inc. and where further details can be found). FCPG, feeding CPG; CMCPG, ciliomotor CPG; LCPG, locomotor CPG; SCPG, shell movement CPG.

Experiments in Syed’s laboratory in Calgary, using cell culture techniques designed to allow transplantation and functional integration of identified neurons into a host nervous system [59], have demonstrated that this is indeed the case. If the soma of RPeD1 was removed and its axons were allowed to degenerate, this significantly modified respiratory behavior. Complete ablation of the RPeD1 cell body and all its branches by direct Pronase injection completely disrupted the respiratory rhythm [55], indicating that RPeD1 is necessary for normal respiratory behavior. Furthermore, following transplantation of a fresh RPeD1 into an animal whose original RPeD1 soma was excised and whose axons had completely degenerated, the transplanted cell reinnervated all its potential targets, allowing restoration of the respiratory rhythm [59,60]. Thus both criteria for a command neuron can be fulfilled by a single neuron, but proving the necessity criterion is often difficult and time-consuming.

RPeD1 is not just an interneuron as it also has direct connections to motor neurons and participates in the respiratory central pattern generator, initiating its activity [56]. Indeed Davis [61] asked “Are command neurons also central pattern generators” in 1978 and RPeD1 certainly fulfils both roles. It is a constituent of the respiratory central pattern generator (RCPG) (Figure 2A), which is distributed across three ganglia within a multi-ganglionic neural network (Figure 2B, Figure 2C) [43,62] and has been reconstructed in vitro [57]. The activity of the RCPG can be terminated by the action of the whole-body withdrawal neuron RPeD11 ([63] and see Figure 2C). However, it is currently unclear whether other neurons can contribute to whole animal withdrawal, which sits at the top of the behavioral hierarchy in Lymnaea [43,62] and presumably in related species such as Helisoma and Planorbis.

3.2 Command Neurons in Simpler Nervous Systems

Classifying a multifunctional, multi-action neuron such as RPeD1 is difficult as it has several functions: it is an interneuron with a command element and also acts as a motor neuron. Perhaps this multifunctionality is forced upon it by the very paucity of neurons within the animal, thus driving the nervous system to develop a suitable behavioral repertoire using very few neurons. This suggests that command-like neurons most usually exist in simpler nervous systems and this view is supported by work on C elegans [36,37,38] and the medicinal leech, Hirudo verbena, whose CNS contains approximately 10,000 neurons [46], rather fewer than that of Lymnaea. Hirudo bilaterally paired neurons descend from the cephalic ganglion and their terminal arbors invade each of the 21 ipsilateral hemiganglia of the nerve cord. The neurons appear necessary and sufficient for intersegmental coordination during crawling [64] and, like RPeD1 in Lymnaea, they are multifunctional: they can command crawling or swimming movements. They are almost certainly command neurons, but detailed experiments such as those in Syed’s’ laboratory remain to be performed. Because of the greater neuronal redundancy in more complex nervous systems single command neurons are much less likely to exist than in simpler nervous systems, so we will now consider the situation in cephalopod mollusks, whose nervous systems are the most complex of all invertebrates [65].

3.3 Cephalopod Molluscs

Cephalopod mollusks include the decapodiform squids such as Loligo spp whose escape systems have been well documented [33,66] and comprise first, second and third-order giant fibres. The two first-order neurons lie in the magnocellular lobe and integrate sensory inputs from the eyes sense organs on the tentacles, etc. These two cells have been assumed to be command interneurons [67]. They fuse in “the midline of the central paliovisceral lobe to form a chiasm where axoplasmic fusion occurs” [33,68,69] allowing synchronization of escape behavior on both sides of the body. When activated they drive 2^nd order giant neurons and in turn the third-order giant neurons, which are motor neurons innervating the circular muscles of the mantle [66,70]. This system needs to be intact to facilitate fast jet propulsion swimming. Thus the first-order giant fibers have been described as command interneurons (e.g. [67]) but most of the support for this view comes from morphological and ultrastructural studies [33,68,71] and most electrophysiological studies have concentrated on the giant synapse between the 2^nd and 3^rd order neurons because of their accessibility and usefulness in understanding synaptic physiology [72]. However, even though the two first-order giant neurons can ensure bilateral, rhythmic contractions of the mantle they should not be considered command neurons because there are parallel motor axons that can generate the strong jet propulsion, associated with escape behavior, sometimes before the giant fiber system is utilized [73] and which presumably interact with them. Thus command-like multifunctional neurons may exist in smaller nervous systems but are much less likely to occur with increasing complexity, which leads to neuronal redundancy and multiple feedback mechanisms. These ideas indicate that there is much less linear neural processing in more advanced nervous systems, when compared with simpler systems with few neurons.

3.4 Command Systems of Neurons in the Vertebrates

Specific vertebrate behavior patterns are now known to be driven by command systems of neurons-see extensive review by Grillner and Manira [4]. For example the locomotor commands are generated by command centers in the mesencephalon and diencephalon, whose activities are modulated by the reticulospinal system, with simultaneous activation of respiration and cardiac output to match increased metabolic needs (see [4] for details). From here information is passed via multiple specialized command lines to specific spinal cord central pattern generators (CPGs) in practically all vertebrates (4). As sensory inputs modulate movements and gait changes the CPGs interact with one another as polymorphic neuronal ensembles as in parallel distributed processing (PDP – see below). However, one pair of vertebrate neurons have previously been classified as command neurons, the Mauthner cells of fish and amphibia [32,74]: a pair of reticulospinal neurons involved in fish escape behaviors (the C-start) that are essential for survival by escaping predation. Further studies have shown that although the firing of Mauthner cells is correlated with escape behaviors, alternative parallel reticulospinal pathways [75] can initiate the behavior when the Mauthner cells are inactivated [76]. Thus, although important, Mauthner cells are part of a reticulospinal pathway and should not be thought of as command neurons. The Mauthner cells have become incorporated into escape behaviors in flying fish and anuran amphibians [75] and it is suspected that “the most fundamental mechanisms for rapid motor control by the Mauthner system may extend even to mammals” [77]. They are of phylogenetic importance [75] and it may well be that something similar happened with the giant fiber system of squids which later became incorporated into the current more complex escape system with many smaller parallel fibers. However, a recent paper by Hecker et al. 2020 [78], suggests that the M cell is essential for rapid escape behavior in zebrafish. However, following ablation of one Mauthner cell body in larval zebrafish this only becomes apparent after the giant axon has fully degenerated over many months. This directly affected survival in predator-prey assays, suggesting that if the Mauthner cells are not integrated into the escape system during development, fast escape becomes impossible, implying that “even complex brains can trust vital functions to individual neurons” [78]. Should the Mauthner cells be considered command neurons, or do they normally work in parallel with other reticulospinal cells? In our view the latter supposition seems most likely, given that two separate command populations of glutamatergic neurons in the brainstem “integrate environmental cues and hence generate flexible swimming movements to match the animal’s behavioral needs” [79] (Figure 1). In this figure animals were chosen to indicate that major differences have occurred both between and within Phyla during evolution, e.g., between the mollusks L. Stagnalis and O. vulgaris; between the arthropods H. gammarus and B. impatiens; between the Zebrafish and higher vertebrates. Animals such as the Zebrafish may represent an evolutionary situation where neural redundancy has not fully overtaken the need for individual command elements such as the Mauthner cells. The environment of each animal no doubt determines many of the specific connections in its nervous system, e.g., the zebrafish depends on the neural connections driving the C-start response for its survival.

4. Parallel Distributed Processing and Modulation of Motor Commands

Parallel distributed processing (PDP), sometimes known as connectionism, was first introduced in the 1980s by Hinton and Anderson [80] and summarised in two edited volumes by and Rumelhart and McClelland in 1986 [81]. In 1992, Kien, McCrohan and Winlow considered the basis of motor programme selection in both vertebrates and invertebrates [82]. They concluded that computation in nervous systems is usually by parallel distributed processing (PDP) rather than by command neurons, often described in reduced preparations, which were unable to behave with normal flexibility. Briefly, they suggested that:

• Small neural networks may be multifunctional when embedded in the CNS. 
• In intact brains, behavioral flexibility is achieved by polymorphic neuronal ensembles of interacting networks to produce a wide range of outputs. 
• Each behavior involves large numbers of neurons.
• Motor output is achieved by PDP not by single cells or nuclei, but from the complex, interacting networks in which these cells are embedded.

These concepts still appear to be relevant today. However, human findings suggest that PDP and much slower serial processing may both occur, particularly when individuals are required to perform two cognitive tasks simultaneously, due to a bottleneck at the decision stage when common circuitry is involved in both tasks [83]. A recent paper [84] suggests that sub-connections of the neocortex fluctuate and are rearranged in the first 200 ms following a visual stimulus in mice, but reach a stable format in 0.5 sec, suggesting sensory coding redundancy near the start of perception. This supports the concept of PDP. As does the Cognitive Evolution Theory (CET) proposed by Yurchenko [85] which considers consciousness to be “a discrete chain of momentary states” and that it has no special neural correlate in the brain but spreads over many brain regions. Given that the different visual stimuli arriving in the brain are divided into four very different components of color, motion, shape and depth [86], this seems an appropriate conclusion and supports the idea of PDP but does not explain consciousness and self-awareness in physiological terms. Interestingly, it has been suggested that people of high intelligence (based on IQ tests) have larger grey matter volumes in the parietal-frontal cortex than those of lesser intelligence. Such data should be treated with skepticism, but higher intelligence in healthy individuals is related to lower values of dendritic density and arborization. In addition, the brains of intelligent people show less neural activity during testing than those of less intelligent people, perhaps using a distributed neural network of “efficiently organized neurons and axons underlying the expression of human intelligence” [87], supporting the PDP concept.

5. Beyond the Connectome

There is more to the physiology of nervous systems than just connectivity and endocrine and neuroendocrine adjustments must be combined with the system. For example, we now realize that action potentials have three functions within nervous systems: communication, modulation and computation [22] using phase ternary, rather than binary computation [20]. The whole point is that nervous systems are likely to be much more complex than circuit diagrams or hormonal and neurosecretory actions can predict, particularly as there is also evidence that brains may function as quantum phase computers [23], exhibiting quantum entanglement [24]. These important concepts will eventually affect our understanding of neuroethology and the organization of animal behavior. Quantum approaches may also start to crack the problem of consciousness, which is unexplained at physiological and connectivity levels. However, philosophers, psychologists and physicists have assumed the importance of quantal organization for some time [88]. Thus there is a basic divide between those neuroscientists “who believe that the way to understanding the brain is through looking at the nerve cells” (i.e., the connectome of various species) and those “physicists who suggest that the laws of quantum mechanics may have an influence on the dynamics of consciousness and the mind” [89]. However, in real neural networks individual neurons appear to act as computational elements and to participate in parallel processing using phase ternary computation [22,23] and in the view of Winlow and Johnson 2023, “the brain acts as a quantum phase computer and would be expected to generate multiple non-classical brain functions in terms of connectivity across a nervous system” [90], presumably interacting with polymorphic parallel distributed neural networks to generate behavior. In analogy, the frequencies of action potentials computed in parallel may be equivalent to central pattern generation. The combination frequencies at any moment will compute as an action, as in the retina [21]. We do not presume that such interactions only occur in the brains of advanced vertebrates, but that they may also occur in cephalopod brains, whose decentralized, distributed nervous systems make them prime candidates for PDP [91]. Whether this would be the case in less complex nervous systems is unknown. However, a possibility to be considered in future studies may be found in Lymnaea and related gastropods, whose RCPG neurons are distributed over several ganglia (Figure 2).

We already know that quantum tunneling occurs through closed voltage-gated channels in biological membranes where small particles can cross energy barriers and may affect the resting membrane potential [92]. Thus some quantal effects are already known in nerve cells and at a psychological level quantal probability has been suggested to provide a framework for understanding reasoning and similarity [93]. Furthermore, Larson [94] suggests the primacy of quantum logic in the cognitive sciences underlying subconscious decision-making. However, until recently there was no neurophysiological evidence for these logical but unproven ideas on non-classical brain functions, although in a recent journal special issue on quantum information applied in neuroscience Georgiev [95,96] suggested that quantum information theory should have significant utility in our understanding of neuroscience and particularly in the precise timing of action potentials [97].

6. Evidence for Non-Classical Human Brain Functions by Quantum Entanglement Between Systems

A non-classical framework for cognitive science was suggested by Horgan and Tienson [98] derived to some extent from the proposals by David Marr in 1982 [99]. However, Marr had provided a useful framework for theorizing about cognition, which had been based on classical artificial intelligence ideas of cognitive science, in terms of three levels of description: cognitive function, algorithm and physical implementation. In 1994 Horgan and Tienson generalized this into a non-classical framework involving cognitive state transitions, mathematical/functional design and physical implementation or realization. Unfortunately they argued against a computational framework for cognitive science while accepting that connectionist networks (PDP) networks were important, which seems contradictory, particularly now that we know that in real neural networks individual neurons act as computational elements [20,22]. However, evidence favoring non-classical brain functions due to quantum entanglement has recently emerged from Trinity College, Dublin in the work of Kerskens and Perez [24], thus supporting our view that the brain uses quantum computation [23].

Quantum entanglement is a perplexing quality of quantal mechanics, suggesting that quanta once entangled can remain so over vast distances. However, entanglement cannot be directly observed and is difficult to detect experimentally, but Rappaport et al [100] indicated that it may be detected indirectly using entanglement witnesses (EW) and this was the approach taken by Kerskens and Perez [24]. They adopted an idea, based on experiments to prove the existence of quantum gravity, postulating experiments in which a known quantum system might interact with an unknown system. If those systems entangle, the unknown must also be a quantum system. They suggested that proton spins of bulk brain water could act as quantum systems. Using nuclear magnetic resonance (NMR) detection they demonstrated, in conscious human volunteers at rest that, in the absence of classical neural signals, what appeared to be heartbeat-evoked potentials were evoked in most parts of the brain using nuclear magnetic resonance (NMR) detection. The volunteers were conscious of these non-classical connections, none detectable with magnetic resonance imaging (MRI). These observations require detailed verification, but suggest that there may be quantum entanglement between systems, not previously observed directly or physiologically identified.

7. Conclusions

The organization of advanced nervous systems is highly complex and motor control resides in command systems of neurons, rather than individual command neurons which appear to occur in simpler nervous systems with restricted numbers of neurons. In these restricted nervous systems, the command elements are often multifunctional and perhaps evolved due to environmental pressures, so that their nervous systems reflect the animal’s physiology, environment and behaviour.

More complex nervous systems have substantial neuronal redundancy and multiple feedback systems to multiple specialised command lines driving central pattern generators (CPGs). Computation of sensory inputs, leading eventually to motor outputs, is most likely by parallel distributed processing achieved from complex interacting networks in which the CPGs are also embedded. This may also be true in some “simpler” nervous systems. The key steps in our understanding of motor command systems are summarised in Table 1.

Table 1 Key concepts in understanding motor command systems.

Although connectivity within nervous systems is vital, this may include more than just a specific connectome because even individual neurons are capable of computation and computational action potentials are considered quantal. Recently quantum entanglement between neural systems in alert humans has been postulated based on physiological evidence indicating non-classical connections between some systems.

Now, we are entering a new and exciting research phase on nervous function, particularly in nonclassical brain connections, as we further investigate the connectome which appears to be underlain by interacting neurons connected by parallel distributed processing.

Author Contributions

WW provided the conceptual framework for this review and wrote the first draft. There were then significant inputs from both ASJ and RF. ASJ helped in rewriting the manuscript.

Funding

The authors received no financial input to support this work.

Competing Interests

We have no conflicts of interest, financial or otherwise.