Single neurons can act as coordinators by selectively combining the activity of different neuronal groups at each moment


Each mental event or voluntary motor act is the result of the simultaneous activity of large groups of neurons in several areas of the brain far from each other.

How do these groups of nerve cells manage to instantly and selectively coordinate their electrical activity?

Discoveries by researchers working with the European Human Brain Project at Universidad Autónoma de Madrid (UAM) and Universidad Politécnica de Madrid, together with collaborators from Jülich Research Centre in Germany, now shed new light on the cellular basis of this process.

The team showed how single neurons, through extremely long branched axonal connections to different areas of the brain can act as “coordinators” by selectively and flexibly combining the activity of different neuronal groups at each moment – similar to the conductor of an orchestra.

The results, which were obtained from mice, were published this week in the Journal of Neuroscience.

Each neuron in the brain has a long, branched extension called an axon, through which it sends electrical signals to thousands of other neurons.

Although they can be hundreds of times thinner than a human hair, axons can be more than a meter long and branch out selectively to reach several points in the brain, and even the spinal cord.

Through its axon, neurons located in areas of the brain distant from each other manage to establish direct contact.

At the contact points, signals pass from one neuron to another through specialized structures called synapses and electrochemical mechanisms mediated by different substances, known as neurotransmitters.

The propagation of signals through the synapses produces simultaneous effects on the neurons that receive contacts from the same axon.

From a functional point of view, synapses can be interpreted as signal filters of variable amplitude and time profile, whose valence (excitation/inhibition) can differ.

The study directed by Prof. Francisco Clascá, from the Department of Anatomy, Histology and Neuroscience of the Faculty of Medicine of the UAM, and carried out on mice, focused on the axons of the neurons that connect the thalamus with the cerebral cortex.

The thalamus is located in the center of the brain and acts as a large communication node between different regions.

The axons of the thalamus neurons innervate all areas of the cerebral cortex in an orderly and selective manner, forming excitatory synapses mediated by the neurotransmitter glutamate.

Many of these axons branch out to selectively innervate two or more areas of the cortex.

This is a 3d reconstruction of a synapse

3D reconstruction of synapses. Image is credited to Javier Rodriguez-Moreno et al.

Using advanced techniques of three-dimensional electron microscopy, performed in collaboration with Prof. Joachim Lübke at Forschungszentrum Jülich in Germany, and tagging of individual axons, the researchers were able to measure and compare the structure of synapses formed by branches of the same thalamic axon in two distant brain areas.

The study revealed important differences, directly related to the intensity and frequency with which the synapses can transmit signals, as well as in the type of cells that the axon contacts in each area.

In a study published by the researchers last year, they had already shown that the signals reach the two areas simultaneously, but produce different effects.

The demonstration that a single neuron, through its branched axon is capable of simultaneously producing different effects in separate areas of the cerebral cortex reveals an unsuspected complexity in the brain’s circuits.

These cells could thus act as “coordinators” by selectively and flexibly combining the activity of different neuronal groups at each moment – similar to the conductor of an orchestra. Knowing more about these cells is important to model the computation performed by the large neuronal networks of the brain and to understand their alteration in brain pathologies.

Neural oscillations, or “Brainwaves,” are fluctuations in activity shared among neuronal populations (evident as extracellular voltage fluctuations; Jia and Kohn, 2011) and were first discovered in the late 19th century in animals (Beck, 1890; Coenen et al., 2014).

The first electroencephalogram (EEG) was performed by Berger in the early 20th century revealing Alpha waves (Berger, 1929) which lead to a volley of research into these waves shortly after. Electromagnetic or EEG synchronization between brain areas indicates functional connectivity between those areas (Ivanitsky et al., 1999).

Even though such oscillations are known to be a component of many cognitive functions such as feature binding, neural communication (Fries, 2005), perception (Gray et al., 1989), and information processing (Gupta et al., 2016), it is still debated whether oscillations contribute to these processes or are merely an epiphenomenon (Koepsell et al., 2010). Various frequency bands of oscillations from very slow (<0.1 Hz) to very fast (600 Hz) have been shown to each be correlated to distinct aspects of mental activity (Stookey et al., 1941; Schnitzler and Gross, 2005; Fingelkurts and Fingelkurts, 2010a), and analysis of the EEG can be used to determine one’s level and potentially state of consciousness (Cvetkovic and Cosic, 2011; Fingelkurts et al., 2013).

Neural oscillations provide a powerful means to encode and transfer information in space and time (Cheong and Levchenko, 2010). They are the most efficient mechanism to transfer such information reciprocally between neural assemblies (Buzsáki and Draguhn, 2004).

They exist at multiple spatial levels from microscopic to macroscopic which can arise from mechanisms within individual neurons as well as interactions between them (Haken, 1996), all of which are a component of the bioelectric structure we describe.

The brainwaves observed on EEG are in fact mesoscopic or macroscopic oscillations (Freeman, 2003). Microscopic oscillations are not as easily detectable. Subthreshold membrane potentials are a major microscopic component of these layers that occur in frequencies observed in an EEG.

Just as action potentials and various types and patterns of synaptic connections serve as a means of information representation, computation, and transmission, subthreshold membrane potential oscillations provide a means for individual neurons to be a part of a collective whole (Fingelkurts et al., 2010).

Such intrinsic single cell oscillations form the basis for frequencies of mesoscopic activity generated by the summed dendritic activity of many neurons within a neural assembly which can be viewed in an EEG (Başar, 2008).

Neuronal assemblies can, in turn, synchronize with other adjacent or distant assemblies to form stronger and more global macroscopic oscillations responsible for the greater neural electromagnetic field (Jirsa and Kelso, 2000).

The emergent characteristic of large-scale bioelectric activity provides a metastable bridge to global coherence needed for an integrated experience (Fingelkurts et al., 2010).

Brains are systems that never reach a truly steady-state, constantly changing in dynamic patterns (Freeman, 2007; Fingelkurts et al., 2009).

A concept of nonlinear dynamics, metastability in regards to the brain describes the local-global harmony of the brain which may be responsible for the emergence of consciousness; distinct functional modules coupled together via neural oscillations while still maintaining their intrinsic, independent behavior (Freeman and Holmes, 2005; Kelso and Tognoli, 2007; Fingelkurts et al., 2013).

There is thus competition in brain regions between the tendency to act autonomously and to cooperate macroscopically with other regions (Bressler and Kelso, 2001; Fingelkurts and Fingelkurts, 2001).

In this metastable mode of functioning, although there is competition between the stability of either tendency, these local and global tendencies can coexist (Kelso and Engstrøm, 2006).

Oscillations may be an optimal metastable mechanism as they provide a low-energy operation for local and distant communication which is lost in action potential signaling in distant axonal connections (Buzsáki and Draguhn, 2004). A relatively large brain with only axonal connections would have severe spatial and metabolic constraints (Knyazev, 2012).

According to the Default Space Theory of Consciousness and other prominent theories on consciousness, consciousness is an emergent phenomenon which arises as the virtual recreation or simulation of the environment and the individual’s relationship to it (Revonsuo, 2006; Fingelkurts et al., 2010; Metzinger, 2013; Jerath et al., 2015a). Metastability, oscillations, and consciousness have been extensively researched as a part of the operational architectonics theory of brain-mind (OA) in an attempt to neurophysiologically explain the integrated experience and mind.

The theory we propose here is in line with the OA argument that the virtual structure of conscious experience corresponds to, or is functionally isomorphic to, the structure or architecture of the brain’s electromagnetic field (Fingelkurts and Fingelkurts, 2001; Fingelkurts et al., 2009).

Functional isomorphism describes two systems as correlating in a way in which functional relations are always preserved regardless of the physical nature of either system (Shapiro, 2000).

For instance, a digital computer can be isomorphic to an analog one if the transitional relations among its physical states mirror those in the analog one (Putnam, 1975).

Thus, whatever the phenomenal constitution of consciousness is at a given time, it will be isomorphic to its neural correlate. OA explains in-depth how any phenomenal state/pattern is reflected appropriately to a neurophysiological state/pattern (Fingelkurts et al., 2007).

A major assumption and basis of this article is a fundamental of OA, that the phenomenal mind is isomorphic to the globally unified electromagnetic field of the brain which consists of a nested hierarchy of oscillatory activity (Fingelkurts et al., 2010).

In this article, we explore a potential implication of an aspect of this architecture that is neglected by most in EEG-based research (Fingelkurts and Fingelkurts, 2010a), that being integrative brain functions arise from the bioelectric architecture of the brain viamultiple oscillations phasically superimposed upon one another based on frequency (Başar et al., 2004; Başar, 2006).

This idea that the true composition of the bioelectric structure consists of a concert of multiple superimposed oscillations is most often neglected as EEG analysis is mostly done by taking different frequency bands in isolation (Fingelkurts and Fingelkurts, 2010a).

Thus, the true bioelectric structure of one brain may be vastly different from another while still having identical averaged spectral band results (Fingelkurts and Fingelkurts, 2010a).

OA has described how at the core of the isomorphism between the neurophysiological organization of the brain and the informational organization of the phenomenal mind lies the “operation,” or the bioelectric processes occurring among the (potentially many) neural assemblies of the brain (Fingelkurts and Fingelkurts, 2005).

Complex operations of synchronized bioelectric activity among distributed neural assemblies, termed operational modules by OA, allow for metastability as the neural assemblies can do their own tasks while still be synchronized with greater and more abstract operations (Fingelkurts et al., 2009).

A potentially infinite nested hierarchy of operational modules, which are at the base level composed of basic operations within neural assemblies, may exist as the simplest modules can become synchronized or abstractly unionized with other modules to form a greater and more abstract module, which can be further unionized with other abstract modules all the way to the most macroscopic level of bioelectric activity proposed to be isomorphic to the integrated experience (Fingelkurts and Fingelkurts, 20052006).

While frequency bands are often identified with distinct functions, some authors have discussed how oscillations of different bands may be grouped into intrinsic layers or “wave-sequences” (Steriade, 2006), or at least superimposed upon other spectrally distinct oscillations (Başar et al., 2004).

In this theoretical article, in contrast to the traditional view that the localization of higher and lower frequency activities are spatially distinct (Luo et al., 2014), we describe an organization of bioelectric cortical neurodynamics modeled as hierarchical “layers” of oscillatory frameworks differentiated by frequency which are not spatially distinct, but coexist in the same brain regions.

The lower layers (low frequency) represent more basic and widespread integrative activity, while the higher layers (high frequency) represent more complex and localized activity.

We thus form a further theoretical understanding on the organization of the global bioelectric architecture, referred to as a unified metastable continuum in OA (Fingelkurts et al., 2009), by describing the superimposition of such layers and its role in such a continuum.

Although divisions and dynamics between these layers may be complex in reality, in basic modeling of such architecture, each layer we describe can be thought of as an independent functional component of this continuum.

The higher layers however are dependent upon the lower ones to be a part of the global architecture as they entrain upon them just as the phenomenal isomorphic counterparts to the higher layers are dependent on the phenomenal isomorphic counterparts of the lower layers.

The fundamental elements of oscillations we see heavily summated (approximately millions of neurons) on an EEG are the ionic current producing membrane potential activities of individual neurons; the dendritic and postsynaptic potentials (Klein and Thorne, 2006).

The activity of individual neurons consists of relatively simple electrical activity, and can thus be considered nonconscious in contrast to the coordinated conscious and unconscious bioelectric activity of neural assemblies which have a phenomenological ontology (Searle, 1992; Fingelkurts et al., 2010).

The phenomenal unity of human experience indicates that there must be some mechanism(s) to unify processes responsible for the many aspects of experience such as the variety of sensory modalities.

We agree with the metastable view that the synchronized operations of several neural assemblies that are integrated EEG spatial-temporal patterns allow for the global functional unity (Honey et al., 2007; Werner, 2009) needed for the integrated experience (Fingelkurts and Fingelkurts, 2005).

While consciousness has been suggested to be quantized (some states more conscious than others; Oizumi et al., 2014), we focus on the phenomenal qualities and contents of human experience in this article.

This article may be seen as a further development of our opinion article on this layered model in which we introduced three separate but highly interactive oscillatory frameworks (Jerath and Crawford, 2015).

We elucidate an updated model here by detailing each of these layers and exploring their nature in different mental states. In the introductory article, we described a base layer of slow oscillations maintained in part by the Default Mode Network (DMN) and cardiorespiratory activity.

The second layer, built upon the first, is a constraining emotional layer powered by the limbic system.

The highest layer consists of higher frequency activity among the elements of the corticothalamic network which creates higher cognitive and perceptual components of mind (Figure 1).

In this updated version of the model, we have separated the infra-slow oscillations and the Delta oscillations into two layers based on their physiological distinctions and focus on the spectral aspects of these layers rather than anatomical locations.

We heavily strengthen our perspective with supporting research and discuss how breathing plays a role in the organization of neural oscillations.

An external file that holds a picture, illustration, etc.
Object name is fnhum-13-00426-g0001.jpg
Introductory model of layered activity. This image comes from our initial article on this theory of which we have significantly improved in the updated version. Panel (A) illustrates the base layer of slow neural oscillations of the Default Mode Network (DMN) and cardiorespiratory activity. This creates a foundation for all other layers of oscillatory activity and is depicted by the blue coloring. Panel (B) shows the second layer of middle-frequency activity largely consisting of limbic activity and is depicted by red coloring. Panel (C) reveals the corticothalamic feedback loops involved in cognitive and consciousness processes and is illustrated with yellow coloring. Panel (D) combines these layers to form the sum of human neural activity consisting of all neural and physiological oscillations. The multi-colored arrows of each person represent the layers interacting with the cardiorespiratory system with the appropriate color for each layer. Previously published in Jerath and Crawford (2015), permission by CC-BY.

In addition to describing the spectral, layered hierarchical framework relative to a global bioelectric architecture, we further extend the oscillatory framework from the brain to neural and non-neural elements of the body.

The relationship between neural oscillations of the brain and activity of the body (largely autonomic) has been explored previously by ourselves and other authors. Coordination and communication between the autonomic bodily system and the brain have been demonstrated in several studies (Walker and Walker, 1983; Basar, 2010).

These links reveal the likely existence of bidirectional oscillatory links between organs of the body and the brain which may allow for the maintenance of survival functions such as body temperature (Achimowicz, 1992; Fingelkurts et al., 2011).

There is also a link between neural oscillations and the immune system of the body (Saphier et al., 1990; Rosenkranz et al., 2003).

In addition, support for the idea that respiration acts as a oscillatory scaffold in the brain is growing (Heck et al., 20162017; Varga and Heck, 2017).

Research into this relationship between the brain and body has not explored how this relationship fits into the global architecture.

We suggest the body fits (largely respiratory elements) into this architecture and may act as an underlying coordinator of bioelectric neural activity.

We also suggest the bioelectric structure of the brain in a sense may be projected to or unified with the sensory receptors of the body.

Although the concept of a hierarchy of brain oscillations across space and time has been previously proposed by notable authors (Freeman, 1987; Lakatos et al., 2005; Knyazev, 2012; Buzsáki et al., 2013; Fingelkurts et al., 2014; Fingelkurts and Fingelkurts, 2017a), we model a hierarchy in a novel way based on frequency by contending that these superimposed spectral layers are isomorphic to superimposed aspects of phenomenal consciousness.

Isomorphism among electromagnetic structure and phenomenal structure has been described (Fingelkurts et al., 2009); however, here we describe an isomorphism between the superimposition of electromagnetic “layers” and the superimposition of various components or “layers” of the phenomenal mind.

The layers we detail in this article have significantly more functionality, detailed operational processes, and blurred spectral borders, however, the modeling of an oscillatory spectral hierarchy which distinguishes groups of oscillatory networks and how the superimpose in relation to the phenomenal mind may further the understanding of the intrinsic and ubiquitous nature of oscillations in relation to psychology.

Human Brain Project


Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.