Scientists studying the properties of the human brain have long known that it works like a powerful computer and is capable, for example, of storing all the information on the Internet.

However, not all factors that determine the “computational abilities” of our brain have yet been discovered.

Another discovery in this area was shared by researchers from the Massachusetts Institute of Technology Institute of Technology. They were the first to record the electrical activity of neurons with an ultra-high level of detail.

It is important to clarify that our brain contains 85-86 billion neurons, and each one functions as an excitable element. It accumulates incoming electrical signals in its body (soma) and, when the voltage reaches a certain limit, it generates a short electrical impulse that is sent to branched processes - dendrites. Note that it is precisely this cumulative approach that allows millions and billions of individual cells to function as a single whole without a common “control center.”

At the ends of the dendrite of each neuron there are membrane projections - spines. The spines of one neuron connect with the spines of another, forming a point of contact - a synapse. It is through them that the nerve impulse is transmitted.

Authors new job decided to compare the “abilities” of dendrites in humans and model animals - rats. They suggested that differences in the functioning of these neural processes were responsible for the brain's processing power and could explain the intellectual superiority of humans over all other species.

Experts explain: each neuron can have up to 50 dendrites, and in humans they are much longer than in rats and most other animals. Therefore, our cerebral cortex is much thicker: it makes up about 75% of the total brain volume (for comparison: in rats - about 30%).

But despite these differences, structural organization This area is similar in rodents and humans: the cerebral cortex consists of six different layers of neurons. In this case, neurons from the fifth layer have the ability to transmit a signal to neurons from the first layer.

But since humans have a much thicker cortex than animals, it turns out that during evolution, neurons had to lengthen their dendrites in order to reach other layers. And the signals themselves travel along such paths longer.

"It's not just that people are smart because we have more neurons and a larger cortex. [Our] neurons act differently," says team leader Mark Harnett.

To take a closer look at how human dendrites work, the researchers used slices of brain tissue from patients with epilepsy. During the operations, volunteers had small (the size of a human fingernail) sections of the anterior temporal lobe removed to gain access to the desired area of ​​the brain.

It is noted that the anterior temporal lobe is responsible for many functions, including linguistic and visual information processing, but removing a tiny part of it does not reduce brain performance. And for neurobiologists, such “living” tissues are unique samples for study.

Once the team received the slices, they were immediately placed in solutions that mimicked cerebrospinal fluid. This made it possible to maintain tissue viability for 48 hours.

The scientists then used an electrophysiological technique called local voltage clamping, which allows them to study the properties of ion channels. There are a lot of the latter in the outer membranes of dendrites, and they are actually responsible for throughput"channel".

Previously, similar experiments were carried out with rodent brain tissue, but the team studied the electrical properties of human dendrites for the first time.

As a result, experts discovered that since human dendrites are longer than rats, the signal coming from a neuron in the first layer to a neuron in the fifth layer is much weaker than the same signal in rodents.

It also turns out that human and rat dendrites have the same number of ion channels, but in our dendrites they have a lower density due to overall dendritic elongation.

This difference may seem to reduce brain performance, but in fact this is not the case. Instead, in order to direct the signal to the desired location, the thousands of synapses on each dendrite must “collectively” determine an “input pattern,” Harnett explains.

Based on the new data, his colleagues developed a detailed biophysical model that shows that changes in ion channel density may explain some of the differences in the electrical activity of human and rat dendrites.

According to Harnett's hypothesis, because of these differences, more parts of the dendrites can influence the strength of the incoming signal, which allows individual neurons in our brain to perform more complex tasks and increase computing power. Brain cells themselves become a kind of mini-computers.

"There is more 'electrical independence' in human neurons, potentially leading to increased computational capabilities of single neurons," he said.

However, there are many other differences in the functioning of the brain of humans and animals, so perhaps the lengthening of dendrites and the associated changes are just one of the advantages that sapiens received during evolution.

In the future, neuroscientists intend to study in more detail the electrical activity of the human brain and find other features responsible for our mental abilities.

Fellow neuroscientists at MIT called the discovery a “remarkable achievement.”

"These are the most detailed measurements of the physiological properties of human neurons to date. These experiments are very complex, even when working with [tissue samples from] mice and rats, so from a technical point of view it is quite surprising that they were able to do this with human tissue." - noted Nelson Spruston from Medical Institute named after Howard Hughes.

Earlier, we recall, the authors of the Vesti. Science” (nauka.vesti.ru) reported that the brains of intellectuals form fewer connections between neurons. Scientists also found in the human brain new type cells and learned how the brain manages to multitask.

For short-term memory to turn into long-term memory, new interneuronal contacts must be formed in the brain, and the formation of such contacts best occurs during sleep activity of nerve cells.

The transformation of short-term memory into long-term memory is called memory consolidation, and neuroscientists are working hard to figure out how and why it happens. Quite a long time ago it was discovered that memory consolidation occurs very well during sleep. That is, in order to remember the textbook you read before the exam, you need to sleep, then the information, as they say, will settle in your head, that is, it will go into long-term storage. There is ample evidence of a connection between sleep and memory. For example, researchers from the University of California, Riverside found that sleeping pills not only normalize sleep, but also improve memory. And their colleagues from the University of California at Los Angeles were able to describe information processes in the brain that are associated with memory consolidation during sleep.

Dendritic spines (colored green) on the surface of neuronal processes. (Photo by skdevitt / Flickr.com.)

Dendritic spines (blue dots) on a neuron. ( Photo The Journal of Cell Biology / Flickr.com.)

That such important process occurs precisely in a dream, it is not surprising: after all, everyone has long known that sleep is just another form of brain activity. It is believed that specific neural impulses, “sleepy” brain waves, are also associated with the fact that our nervous system is engaged in sorting the information received during the day, while external signals do not interfere. But for a long time biologists were unable to figure out how neurons behave and what cellular and molecular mechanisms are involved.

To find out what happens to neurons during memory consolidation, Wen-Biao Gan ( Wen-Biao Gan) and his colleagues from New York University created a genetically modified mouse in which a fluorescent protein was synthesized in the neurons of the motor cortex. With its help, it was possible to observe changes in nerve cells, for example, where and when dendritic spines, special outgrowths on the dendritic processes of nerve cells, are formed. The appearance of a spine indicates that in this place the neuron is ready to create contact with another neuron, in other words, the spine precedes the synapse. Thanks to synapses, neural circuits are formed that are needed to remember information. When we learn to ride a bicycle, for example, we develop new neural circuits in our brains that arise in response to the need to coordinate muscle efforts in a new way. Then, when we get back on the bike, these neural circuits are turned on again - unless, of course, they have disintegrated for some reason, if the synapses between neurons have not disappeared. Returning to dendritic spines, we can say that they indicate the neuron’s response to new information and about the readiness to remember it.

Actually, in the experiment, the mice were also given something like a bicycle: the animals had to maintain balance on a spinning stick, which was spinning faster and faster. Over time, the mice remembered what to do and no longer fell off it. At the same time, the same dendritic outgrowths appeared in the neurons of the motor cortex - the cells understood that the new stimulus was important for the body and were preparing to form new circuits. Then the researchers changed the experimental conditions: the mice were trained on a rotating stick for one hour, but then some animals were sent to sleep for seven hours, while others had to stay awake for the same amount of time. It turned out that in those mice that were allowed to sleep, dendritic spines grew more actively. In other words, sleep helped nerve cells tune in to remember new information.

Moreover, the nature of the appearance of dendritic outgrowths depended on what kind of exercise needed to be performed. For example, if a mouse had to walk along a rotating stick in one direction, then spines appeared on some dendrites, and if it was necessary to walk in the other direction, then spines appeared on other dendrites. That is, the cellular morphology of neural processes depended on what kind of information needed to be processed.

Finally, neuroscientists were able to show that the cells of the motor cortex, on which exercise performance depended, were activated during the slow-wave phase of sleep. Such activation during sleep was important for the formation of the notorious spines: if the “sleepy” activity of the cells was suppressed, then the spines did not form. It was as if the brain was replaying for itself what it should have recently done while awake - replaying it to remember better.

The result was the following scheme: neurons receive some stimulus or perform some procedure during wakefulness, then during sleep these cells are activated again, and such re-activation stimulates cellular rearrangements that contribute to long-term memory of the stimulus. Neuroscientists have long assumed that this is exactly what is happening, but now they have been able to obtain experimental confirmation, and not on just any fruit flies, but on the brains of mammals. Although, of course, now scientists need to find out what molecular processes are involved here, what genes and proteins control the increase in dendritic spines during sleep, what signaling pathways work here, etc.

Speaking of fruit flies: a few years ago, researchers from Washington University in St. Louis and the University of Wisconsin in Madison conducted similar experiments with fruit flies, and then the results showed the same thing - that sleep is necessary for memory consolidation. However, at the same time, neuroscientists observed the clearing of the Drosophila brain from synapses, that is, something like editing of nerve circuits was happening, clearing neurons of unnecessary connections that would take resources away from necessary contacts. Most likely, such elimination of unnecessary synapses is not a specific process characteristic only of insects (or arthropods, or invertebrates), and in the brain of higher animals at the moment of “sleepy” consolidation of memory, along with the formation of new synapses, the breaking of old ones also occurs - we just have to see this in the experiment.

Many nerve cells are like bushes or trees: their output process, the axon, is the thin root of this tree, all the other numerous processes are dendrites. Dendrites usually extend from the cell body in the form of thick trunks, which are then divided into several thinner branches, which, in turn, into even thinner ones, etc. The length of dendrites is tens of times greater than the diameter of nerve cells, and the thickness of the terminal The branches are very small - can be fractions of a micrometer. The question of what role dendrites play in the functioning of nerve cells has not yet been completely resolved and, most likely, their role is different in different neurons. In particular, in some cells the dendritic membrane is inexcitable and can transmit signals only electrotonically, like a passive cable, while in others, the dendrites are capable of conducting AP. Now we will consider only those properties of dendrites that are associated with their geometry.

Let us first consider those cells whose dendrites are nonexcitable. In this case, the “dendrite problem” is as follows. Synaptic endings occur at the most different parts dendritic tree. Let's take a synapse acting on a branch that is maximally distant from the cell body. In this case, the conditions for transmitting an electrical signal seem to be very unfavorable. Indeed, in a thin branch the damping constant is high, and at its end the branch “flows” into a wider section of the dendrite, which “short-circuits” it. In such short-circuited cables, the potential drops especially strongly , However, in the case of dendrites, the “shorting” is incomplete and the potential at the end of the branch does not drop to zero. In the next segment of the dendrite, the conditions for signal transmission are also unfavorable, since at its end there is also a thicker dendritic trunk, etc. “In this regard, the idea arose that synapses located on distant branches make a very small contribution to the change potential of the cell body, hundreds of times less than the same synapses on the cell body. It turns out that synapses on terminal dendritic branches are useless, that this is a “mistake of nature.”

One of the options for solving the “dendrite problem” is that many synapses can be placed on thin terminal branches, then the joint action of these synapses will be noticeable in the cell body. But for this it is necessary that all these synapses work more or less simultaneously.

All the above arguments have long been of a qualitative nature. In 1965, the Theoretical Department of the Institute of Biophysics of the USSR Academy of Sciences developed a method for quantitatively assessing the efficiency of synapses for nerve cells of any shape and calculated this efficiency for motor neurons, pyramidal cells of the cortex and cerebellar cells. It turned out that the efficiency of dendritic synapses is only 3-5 times lower than that of synapses located on the neuron body. What explains this? Why did the efficiency of remote dendritic synapses turn out to be quite high? The smaller the cell, the higher its input resistance, the greater the potential shift the synapse creates. In thin dendritic branches, distant from the cell body, the input resistance turned out to be large, so synapses can create potential shifts in these branches that are tens of times greater than in the body of neurons. And although this potential shift does indeed attenuate greatly when propagating toward the body, its large magnitude largely compensates for the attenuation. Thus, dendritic synapses turned out to be not a mistake of nature at all.

Now let’s consider those neurons whose dendrites have an excitable membrane capable of generating APs. In such neurons, the high efficiency of the synapse on a thin branch can lead to the fact that only a few synapses will lead to membrane potential to the threshold and will cause an PD in this branch, which will begin to spread to the cell body.

His further fate depends on the properties of the branching nodes through which it must pass on the way to the cell body, i.e., on the geometry of the dendrite. This type of cell works like a complex logic circuit. An example of such a cell was shown in Fig. 45; this cell detects unidirectional movements of the stimulus. Cells with more complex shape dendrites can work as quite cunning computing machines. “Such a system is similar to a voting system with a large number participants who have an unequal number of votes... Final result of course depends on total number votes cast "for" or "against", but he is not to a lesser extent depends on who exactly and with which partner votes,” wrote employees of the Theoretical Department of the Institute of Biophysics of the USSR Academy of Sciences in 1966.

The dendrites of many neurons have special formations, the so-called spines. These are mushroom-like structures consisting of a head on a thin stalk, more commonly called a spine neck. The spine is a protrusion of the cell membrane, and a terminal from another neuron approaches its head and forms a chemical synapse on it.

Why spines are needed is unknown. The number of hypotheses about their functions is enormous. Let's see what can be said about the possible functions of spines based on geometric considerations. In this case, we consider two options: the membrane of the head of the spine is inexcitable; the spine head membrane is capable of generating APs.

Let the spine not be excited. Its thin neck has high resistance. As a result, a large postsynaptic potential will arise in the head, but a noticeable part of it will be lost in the neck. The spine will act as a thin dendritic branch. But why do you need such a device? Why shouldn't the synapse be located directly on the dendrite?

One way inhibitory synapses work is to reduce the input resistance of the neuron. But excitatory synapses also open ion channels and reduce input resistance! Because of this, excitatory synapses also interfere with each other. Such interference will be especially strong on thin dendrites, which have a very high input resistance, so that the activation of several synapses will cause a noticeable decrease in it. The spines should significantly reduce the mutual influence of neighboring synapses, which in this case are separated from each other by necks with high resistance. Calculations have confirmed that although spiny synapses are each individually less effective than synapses located directly on the dendrite, when working together the effect is noticeably higher.

If the spine membrane is excitable, then it can work as an amplifier of synaptic transmission. Due to the thinness of the neck, the input resistance of the spine is very high and one synapse can cause an AP in the head, which will send a much stronger electric current into the dendrite than the current of the synapse. It is interesting that with this mode of operation of the spine, there should be an optimal resistance of its neck. It should not be too small - then a noticeable part of the synaptic current will flow into the dendritic branch, the potential shift on the membrane of the spine head will not reach the threshold value and AP will not occur there. But, on the other hand, the resistance of the spine neck should not be too large, otherwise too weak a current will flow from the spine head into the dendrite and no increase in the synaptic current will result. Recently, work has appeared showing that the geometric structure of real spines is close to that which, according to theoretical calculations, is optimal.

So far we have talked about the shape of fibers and cells or even microstructures of cells - spines. Let's now look at the geometry of cellular associations.

Nerve cells transplanted into the adult brain established correct contacts with “local” ones and joined the overall work.

Nerve cells, as we now know, although they are restored, are still not as fast as we would like. On the other hand, it is now possible to grow the most different types cells, including neurons.

It would be good if, in the case of a disease accompanied by massive death of neurons (such as a stroke or Parkinson’s or Alzheimer’s syndromes), it would be possible to transplant new, fresh and healthy cells instead of dead ones - just like replacing burnt electrical wiring or a damaged part of a microcircuit. However, neurons are known to be connected to each other by many contacts and participate in a variety of nervous processes, and therefore, if we want to transplant something into the adult brain, we first need to answer the question: will new elements be able to find their place in it, integrate into the nerve circuits?

Two years ago, we wrote about the experiments of researchers from the University of Luxembourg, who transplanted neuronal precursor cells into mice into the cerebral cortex and hippocampus (one of the main memory centers) - according to the authors of that work, the cells successfully matured in a new place, established contacts with nervous chains, and . That is, in principle, the brain accepts the transplanted neurons; but to understand whether they are useful, whether they participate in information processes, new experiments were required.

And now in Nature article by Suzanne Faulkner ( Susanne Falkner) and her colleagues from the Max Planck Institute for Neurobiology and the Ludwig Maximilian University of Munich, who found that if neurons are transplanted into the visual cortex, they not only integrate correctly into the neural circuits, but also improve vision.

The visual cortex, compared to other areas of the brain, has been particularly well studied; we know about its neurons, when and why they turn on and off, and with what other areas of the brain they are connected. In the experiment, a fragment of the visual cortex was removed from mice, and in its place a piece of the brain cortex taken from the embryo was transplanted, and then, using a special microscopic technique, individual cells were observed.

Within a month, according to the authors of the work, the transplanted “protoneurons” were normally transformed into mature neurons, going through the same stages that maturing nerve cells usually go through. (In particular, in those transplanted over time, the number of dendritic spines also decreased in the same way - areas on the membrane of a neuron where a synapse can form, contact with the process of another neuron; it is believed that a decrease in the number of spines helps to better organize information flows, helps nerve cells not get confused in a huge number impulses entering the brain.)

However, neuroscientists wanted more: their goal was to see that each individual cell, after transplantation, not only turned into a normal neuron, but also made the correct connections with others. In other words, here it was necessary to analyze the connectome of the transplanted fragment: the direction of interneuron connections that went to other areas of the cortex, and their strength.

It turned out that in the operated mice the situation is the same as in ordinary mice that did not receive anything transplanted. In other words, the “incoming” cells not only established contacts with those with whom they needed to, but also the strength of such contacts was the same as it should be (somewhere weaker, somewhere stronger, depending on with whom the given area of ​​the cortex exchanges information). There were some discrepancies with the “original”, some neurons established synapses with the wrong ones, but the reason here, obviously, was that a piece was taken for transplantation that did not exactly match the one that was taken out of the brain. And next time it is quite possible to avoid incorrect connections if you carry out the entire procedure more accurately.

Finally, the last test - a test of functionality - the transplanted cells also passed successfully. The mice were periodically shown certain patterns of stripes, and gradually the new cells learned to distinguish some patterns from others: they responded to some more strongly than to others. That is, over time, the tincture occurred, the training of nerve cells, which, as we remember, were not in the brain from the very beginning.

So, thanks to the fact that the authors of the work monitored the fate of individual neurons, they were eventually able to establish quite reliably that the transplanted cells not only integrate into the system of already formed nerve chains, but also begin to work quite successfully. (Which is especially curious, since it is about the visual cortex that it is believed that it is not prone to restructuring.)

In the future, researchers are going to find out how neurons obtained in a different way will behave (that is, not taken from the brain of an embryo, but, for example, grown after reprogramming skin cells through the stage of induced stem cells), and whether such patches can be used to treat natural brain damage – for example, in case of physical injury or stroke.