Human brain theory

14.3 The generation of motor clock signals

Movements of the simplest animals were initially triggered by receptor signals that reached the muscle plate via axons and led to contraction. This was certainly already organised in this way in the unsegmented animals. There is no question that receptors could also act on muscle, which rather detected data for the life support system, such as oxygen saturation or a lack of energy. Then such receptors also activated the motor function.

With the development of segmented animals, motor reactions were initially limited to the segment whose receptors were activated. Only with the development of a cross-segmental nervous system was it possible to activate the muscles of neighbouring segments.

The organisation of the cord ladder nervous system with its sensory and motor ganglia and its midline centres, as well as the development of a simple primordial brain, created the basis for a central control of motor activities that has been maintained in vertebrates. These foundations now need to be deciphered. We focus here on vertebrates, whose body structure is clearly more complex.

 

Movements of vertebrates often contain periodic components. The flight of a bird, the running of a crocodile, a dog or a horse, but also the swimming of a fish or the slithering of a snake consist of sequences of movements that are constantly repeated.

The signals that control such periodic movements must necessarily also be periodic.

And since every start of movement requires neuronal activation in the system, there must be the possibility in the nervous system of vertebrates to activate movement signals somehow.

Decades ago, when first considering a brain theory, the author noticed the magnocellular system in the brain. Later he introduced the concept of mean systems.

Signal averages - formed by different receptors and average neurons - form the basis of all motor control, according to the author. Since movements initially serve primarily to search for food, signal averages must be the output signals from which movement signals ultimately arise. Receptors that analyse the lack of energy in the system, for example, could flow into the signal averages for movement generation. Similarly, signals from the circadian system, which increase in strength at dawn, could flow into the signal chain to trigger movement. Olfactory signals reveal prey and can be useful for movement activation. But visual signals triggered by prey could also initiate movements, as could acoustic signals, not forgetting the lateral line system, which represents a long-distance sense of location.

From these signal averages, the nervous system of vertebrates must first generate periodic signal sequences.

If you take a closer look at the movement sequences of vertebrates, you will see that there are usually two parts of the movement that are opposed to each other. When a four-legged vertebrate runs, the movements of the left extremities are opposite to the movements of the right extremities. But the extremities of one side also move in opposite directions. If the right front leg swings forward, the right hind leg swings backward. On the left side of the body it is exactly the opposite. The reason for this is the bilateral body structure.

Therefore, a motion control system needs two signal sequences that run in opposite directions and are also periodic. Comparable to a sine curve for the one movement component and a cosine function for the complementary movement component. However, the sine curve would first have to be translated into a frequency-modulated sequence of action potentials, as would the cosine curve. Both would be shifted in relation to each other. Physicists would say that both movement signals are out of phase with each other.

Neurologists usually locate the generation of such periodic and partly phase-shifted neuronal signal sequences in so-called clock generators. The term clock generator implies that there is a neuronal centre that generates periodic signal sequences as output. This is a gross simplification. Up to now, we obviously do not know how these signal sequences are generated and think of a kind of black box that is given the name clock generator. And the proof that there are neurons at a certain location in the brain or in the spinal cord whose activity resembles a clock for movement control is not yet proof that this location contains the presumed clock generator, because neuronal signals are distributed quite widely in the nervous system of projection neurons, so that the same signal will be found at several locations in the brain or the spinal cord.

Regardless, the circadian clock generator has indeed been located in the nucleus suprachiasmaticus.

In this brain theory, it is shown below that the generation of periodic and phase-shifted signal sequences for movement control is made possible by the interaction of several different modules of the brain (or spinal cord). The starting point is the mean value signals that serve to control life processes.

So let's assume that because of a detected lack of energy or some other reason, the life support system generates an impulse in the responsible mean centre, produced by a wide variety of receptors, indicating that it would be time to initiate movements for the purpose of catching prey. One impulse is sufficient, for example a firing rate of 300 Hz over the period of one second. No continuous signal is necessary, as will be shown below.

From this short initial oscillation, i.e. the start signal, a permanent periodic oscillation must now be generated. This is physically only possible through feedback.

We visualise that this start signal propagates along a fixed path. It is generated by a mean nucleus. This projects excitingly into class 1 neurons of the sensory nucleus of the same floor. From there, the signal is transmitted via interneurons to class 4 projection neurons, on whose axons it ascends headward until it reaches the cortical floor. Here, it changes the side of the body beforehand in the crossing floor. In the sensory cortex, it is switched to projection neurons of class 3, which project to the motor side. Their target is class 5 projection neurons, which project downwards, changing sides at the level of the crossing floor to the original side of the body. The signal continues to descend until it reaches the nucleus ruber, for example.

But the cortical floor not only projects descending to the nucleus ruber, but also into the substantia nigra pars compacta. Since the crossing floor is also passed through here, the signals land again on the original side of the body. There, this signal - which, because of the greater distance and the low degree of myelination, advances somewhat more slowly and is therefore time-delayed - is switched to dopamine and sent back towards the cortex, where, however, it ends up in the striosomes of the striatum and excites them. Again, it passes through the crossing floor as it ascends. The striosomes switch it to the inhibitory transmitter GABA and send it back towards the nucleus ruber via the crossing floor. Here, too, the signal suffers a further time delay due to the finite speed of propagation.

 

 In the nucleus ruber, this time-delayed basal ganglia signal now inhibits the original, excitatory mean value signal (start signal) in a point-to-point projection. The inhibition is not total, but only relative, because the mean value signal originates from a mean value neuron, which is difficult to inhibit. Therefore, its firing rate is only slightly reduced. In this phase, the firing rate of the signal is strictly monotonically decreasing. In terms of time, the resulting signal decreases in strength.

Now the coupling of the two halves of the body via the spinocerebellum comes into play.

We call the previous signal S_i , because it is assigned to the ipsilateral half of the body. From it, the spinocerebellum generates a second, phase-shifted signal S_k in the contralateral half of the body.

For this purpose, the signal S_i is transmitted from the ipsilateral nucleus ruber to the nucleus olivaris. This is the lateral change nucleus of the initial floor of the brain. It projects into the contralateral spinocelebellum. In this, the signal is inverted, i.e. its signal strength is reversed. If the input is strong, the output is weak; if the input is weak, the output is strong. This is because the nucleus olivaris first switches the signal S_i to Gaba and this inhibits a permanently excited output neuron in the cerebellar nucleus. This reverses the signal strength, i.e. inverts it.

While the signal S_i on the ipsilateral side becomes weaker and weaker due to the basal ganglia inhibition, the signal strength of S_k on the contralateral side increases. When S_i becomes zero, S_k reaches its maximum.

Now this process would have to lead to a stable phase, S_i would remain eternally equal to zero, while S_k would remain at the maximum value. But the basal ganglia act on both halves of the body.

The contralateral signal S_k of the nucleus ruber of the opposite side now also ascends to the cortex and changes from the sensory to the motor side. There it descends not only to the nucleus ruber, but also to the substantia nigra pars compacta. This again projects excitatory into the striosomes, but this time on the contralateral side. The striosomes again project inhibitory into the nucleus ruber, which this time is also on the contralateral side. In a point-to-point mapping, the inhibitory basal ganglia signal again arrives exactly at the neuron that contained the signal S_k and inhibits it. As a result, the firing rate of the signal S_k suddenly becomes smaller than the original maximum value. In this phase, the signal strength of S_k thus becomes smaller until the value zero is reached. The signal strength is strictly monotonically decreasing.

At the same time, the signal S_k from the contralateral nucleus ruber reaches the olive again and via it the ipsilateral spinocerebellum, which again generates the inverted signal S_i from it. While S_k becomes weaker, the signal strength of S_i increases again up to the maximum value, which is limited by the refractory period. Now everything repeats itself.

The end result is two continuous oscillations whose firing rate exhibits a frequency modulation and which are out of phase with each other.

If the ipsilateral oscillation reaches its maximum, the contralateral oscillation has a zero rate of fire. If the rate of fire decreases ipsilaterally, the rate of fire increases contralaterally. If the rate of fire falls on the ipsilateral side, it increases on the contralateral side.

The interaction of the excitatory back projection from the contralateral cerebellum and the inhibitory back projection from the ipsilateral striosomes of the basal ganglia produces two periodic oscillations with respect to the firing rate, which are out of phase and with which, for example, movements on the two halves of the body could be produced out of phase. On the same side of the body, the front extremities can be controlled out of phase with the rear extremities.

Here it becomes clear that there is no motor clock centre in the classical sense. The vibration-generating substructures are distributed among the nucleus ruber, the nucleus olivaris, the spinocerebellum, the cerebellum nucleus, the cortex, the substantia nigra pars compacta and the striosomes, and this both on the ipsilateral and on the contralateral side.

Now we only have to show how the two phase-shifted oscillations are translated into movements of the body. Structures of the vertebrate brain that we already know serve this purpose: Divergence modules (or in older monographs of mine also called divergence grids).

It must be remembered that to start these two phase-shifted, periodic oscillations, only a short start impulse from a mean value kernel is required, after which both oscillations remain stable and must be stopped by other mean value signals if necessary. This explains, for example, why horses can run away. Once set in motion, they flee hastily without stopping of their own accord. Only a very strong stop signal can inhibit their flight.

This shows that answering a question usually raises new questions. In this case, it must be explained how a movement that has been set in motion can be interrupted again. This explanation may be found by others.

We will refer to the two oscillations S_i and S_k in this monograph as motor clock signals. They are capable of generating movements in divergence modules of all kinds. This will be shown in the next chapter.

Before we describe the generation of movements, however, we must return to the spinocerebellum. The spinocerebellum not only projects to the nucleus ruber and via this downwards to the motor neurons of the spinal cord, but also headwards to the cortex. Here, however, a special feature should be noted. The ascending, cortical projection switches to the other side of the body. The crossing of the upper cerebellar peduncle serves this purpose .

Therefore, when cortical signals from one half of the cortex travel via the nucleus ruber and the olive to the cerebellum of the opposite side, they land from there via the junction of the superior cerebellar peduncle back on the original half of the cortex. However, since the cerebellum signals are the inverted cortex signals, each half of the cortex generally receives both types of signals: the direct ones from the thalamus and the inverse ones from the spiocerebellum.

We assume that the direct signals in the cortex are the original ones, they existed already before the formation of the cerebellum.

The newer cerebellum signals were processed by new cortex neurons, which, as expected, arranged themselves above the older neuron layer of the direct signals.

Thus, in the (motor) cortex, the frontal cortex, there is apparently a structure of two input layers: the lower direct input layer with output from the spinal cord, for example, and the outer, indirect input layer with cerebellum output.

The lower input layer contains only On-type neurons. For example, as muscle tension increases, their firing rate increases.

The upper input layer contains only Off-type neurons. With increasing muscle tension, the firing rate then decreases because this signal was derived from the On type by signal inversion in the cerebellum. On signals become Off signals through signal inversion.

In addition to the On signals of the muscle tension receptors (tendon receptors) and other receptor types, the motor cortex also contains the mean signals and the motor clock signals derived from them. Here, too, there is the direct type, the on-type, which corresponds to the signal S_i and which ends at the input neurons of the lower input layer. The signal S_k , which is inverted from the cerebellum and arrives in the same half of the cortex via the junction of the upper cerebellar peduncle, is of the Off type and ends in the upper input layer of the cortex. Both signals are inverse and out of phase with each other. We assume a spatial well-ordering in the cortex layer. Input neurons with the same signal origin lie exactly on top of each other, the ON neuron at the bottom, the Off neuron at the top. In between are uncounted interneurons and equally uncounted output neurons in the form of small or larger pyramidal cells.

This means that the output of the cerebellum is treated as a new modality that is placed in a new layer of neuron class 4 above the previous cortical neuron layer 4. Something similar is observed in the neuronal layers of the visual thalamus or primary visual cortex. It seems to be a basic principle of vertebrate brain organisation to place new modalities on an equal footing with the old, previous modalities.

 

 

14 .3 Movement generation at the simple joint

 

In the previous chapter, it was explained how a start signal from the mean value system for life support - triggered, for example, by a lack of energy - produced two mutually phase-shifted and periodic continuous oscillations that appeared suitable for motor control of the body.

Now the reader may expect some new, hitherto unknown neuronal circuit that could generate movements from these two motor clock signals. Some will expect a complicated, possibly quite opaque circuit in the brain. But this is not the case!

The circuits that generate movements from two motor clock signals have been present in the vertebrate brain for ages and are quite simple. One way to generate movements from clock signals is to use ordinary divergence modules (also called divergence grids in my earlier monographs).

Divergent modules and convergent modules differ in terms of input and output. Divergent modules are located in the ascending sensory branch of the cord ladder system or on the sensory half of the neural tube.

Convergence modules, on the other hand, are found on the motor side.

Both types of modules arose from the need to protect neuronal circuits against the failure of individual neurons. For this purpose, neuronal signals were distributed among several neurons. This happens in divergence modules. In convergence modules, the signal divergence was reversed. In both cases, there was the option of initially arranging these reserve neurons as vertical neuron columns. This had the advantage that the topological arrangement was preserved in the area. Thus, the neurons in the visual cortex, but also in the corpus geniculatum laterale, are well-ordered in such neuron columns, which are called ocular dominance columns.

These simplest divergence modules are also the ones that first emerged in the course of evolution and are still found today in reptiles and birds. They are divergence modules with vertical signal propagation.

It should be noted that the input for such divergence modules with vertical signal propagation basically consists of a pair of mutually inverse signals that are fed into the neuron column at the top and bottom, respectively. If an original quantity, for example brightness, produces a firing rate in a receptor that increases with the strength of the original quantity, this signal is referred to as the on signal. The signal inverse to this decreases in strength with the firing rate when the strength of the primal quantity increases. Both signals are inverse to each other.

In the divergence module with vertical signal propagation, the neurons of the on-signal form a thin neuron layer. The neurons of the off-signal form a second thin neuron layer. Between the two input layers there is a relatively thick output layer, which could also be understood as a stack of thin output layers. With increasing evolution, the number of stacks becomes larger; initially, there was only one thin outpup layer.

Whether the on-layer is located at the top and the off-layer at the bottom depends on which type of receptor was present earlier in evolution.

This is where the cerebellum comes into play. After the development of the cerebellum, it was no longer necessary for each type of signal to form an inverse type of signal. In the visual area - which developed very early - such a splitting of signals into on-types and off-types could develop. This did not happen with many types of receptors that concerned motor functions. There are no "inverse" muscle spindles whose signal strength decreases with increasing muscle tension. There are no "inverse" tendon organs whose firing rate increases with decreasing muscle tension. The task of forming an inverse signal to a motor-effective signal was taken over by the cerebellum. In both the vestibulocerebellum and the spinicerebellum, signal inversion was originally the main task.

The cerebellum output found its way - in the course of evolution and certainly quite early - to the cortex, i.e. to the top floor of the early cord ladder system or to the cortical floor of the neuroanatomical system. Since it also changed sides, it arrived in the cortex together with the original signals. The original signals already existed before the development of the cerebellum, so they were the evolutionarily older signals. They therefore formed the lower input layer in the frontal lobe. The cerebellum signals, which arose through signal inversion, formed the upper input layer in the frontal cortex. They were arranged in a topologically well-ordered way so that the corresponding input neuron with the off-signal from the cerebellum was located exactly above each input neuron that received a certain on-signal. Between them were the output neurons, which received the input of the output neurons above or below them via a myriad of interneurons. They formed the output column, so to speak. The signal propagation was basically vertical, hence the name divergence modules with vertical signal propagation.

So while the on-signal from the lower input layer propagated vertically upwards, the off-signal from the upper input layer fought its way downwards. All output neurons of the vertical neuron column between the two input neurons with the signals inverted to each other received both signal components because they propagated vertically via interneurons. We can combine these neurons into an output column. This is the typical organisation in the vertebrate cortex. There were no contacts with the neighbouring output neurons of the neighbouring columns, the distance was too great and the neurons were not signal-compatible - they belonged to other muscle groups.

Now, if the joint in question did not produce any signals because both muscles were slack, then there was no output at all in the output column in question.

Now, signal propagation in the divergence lattice is not lossless. On its way from the two input neurons to the output neurons of the cell column, a non-linear signal attenuation occurred. The neuronal excitation of a nerve cell ultimately resulted from the difference in concentration of the differently charged ions. Since there was a drop in concentration in the cell membrane due to constantly open ion channels and the movement of the ions was hindered by the electrical resistance of the cell plasma, the signal attenuation increased with increasing distance from the input source. We establish here a quadratic dependence on the distance.

But how could such a divergence grid move a joint?

 

Here we make a recourse to the mean value system. As described in the previous section, this generates two mutually phase-shifted clock signals, one of the on-type and one of the off-type. We now recall the back-projection of all mean value systems into activation neurons of neuron class 1.

We also find these activation neurons in the cortex. We assume that the motor clock signals - formed from motorically effective mean value signals - end in the cortical target area where the motor control signals of the muscles are also located. Therefore, these clock signals - among others - also end up near the neuronal column that is assigned to our joint.

The signals from the tendon organs of this muscle travel to the sensory centre and contact a class 4 neuron via the axons of the receptors. This neuron transports the signal to the sensory part of the frontal lobe with the interposition of further class 4 neurons. A class 3 neuron takes over the excitation and projects into a class 5 motor projection neuron, which again projects via further interconnected class 5 neurons to a motor neuron of the floor in which the siglal-producing muscle is located.

If the muscle is flaccid, the signalling pathway via the described neuron chain exists, but it is signal-free. This changes when the neuronal clock signal, the origin of which we have described, excites a cortical activation neuron of class 1 from the associated mean centre. Then this excitation is transmitted by interneurons to the neuron of class 4 or class 5, which lies in the signalling pathway of the muscle concerned. Via the descending projection from the cortex to the motor neuron, the muscle is excited and therefore contracts.

 

We assume that the on-tact signal is located below the on-inpuch layer. The associated activation neuron has a larger axon radius and distributes its excitation into the surrounding area.

The associated off-beat signal lands above the on-input layer, where there was the original layer of class 1 activation neurons.

Thus, the on signal of the motor clock signal is fed to the neuron column of the joint from below, while the motor off signal is fed from above. The on signal excites one of the joint's contraction muscles, the off signal excites the opposite. And since the firing rate of the motor on-stroke signal resembles a sinusoid, as being periodic, the one muscle of the joint is contracted periodically, while the motor off-stroke is contracted out of phase each time, since it is excited inversely by the cerebellum. This triggers a periodic movement in the joint.

A periodic movement of a joint (with one degree of freedom) is thus generated by the motor clock signal formed from average values.

In the case described, we had located the motor output neurons in the frontal cortex. However, this is not absolutely necessary.

Likewise, we could have located these neuron columns in the torus semicircularis. This system is connected to the basal ganglia system on the one hand and to the spinocerebellum on the other, and both were involved in the mean system. The formation of the neuronal columns in the torus semicircularis could therefore have taken place in an earlier evolutionary period, and later the frontal contex took over more specific tasks. The reason for this was the splitting of the modalities and segments above the torus semicircularis, so that certain joints received their own spatial representations in the frontal cortex and could thus be controlled separately. This expanded the possible range of movement of the vertebrate.

In any case, it could be explained why an animal whose cortex has been removed is still capable of higher movements, such as walking, if it is supported by suitable measures so that the legs can still swing freely. Here, motor control takes place in the deeper brain nuclei.