(ORDO NEWS) — According to the proposed model, the algorithm for directing axons to target neurons is encoded in the very process of brain cell division. Its essence is in the search for “cousins” according to the “family tree”.
Researchers at ETH Zurich are claiming a partial solution to one of biology’s major problems. They developed a general model of axonal guidance, in which brain cells communicate with each other into complex neural structures
The brain of a mammal, including humans, is the most complex biological structure, consisting of tens of billions of cells (neurons). With their processes (axons) they connect into neural networks, thanks to which the brain can remember, feel and think.
In this case, each axon, according to some principle , chooses which neuron to connect with. It could be a neuron nearby, or it could be a neuron on the other side of the brain.
Exactly how axons do this is still unknown. At the local level, the sprouting of axons is well described: they are drawn to certain molecular marks, first to one, then to another, third, and so on.
But what globally causes the axon to sort out labels and pave the way, sometimes a considerable one, among the mass of other neurons? How does each axon find its single target neuron?
Complicating the problem is that the formation of neural connections begins in the womb. This means that this process is encoded at the genetic level. However, a complete map of neural connections ( the connectome ) simply cannot fit into DNA.
For example, the information capacity of the mouse genome in a germ cell does not exceed one gigabyte, while the most rough and approximate mouse brain connectome requires 10 terabytes of storage.
Swiss scientists suggested that the genes encoded not by the connect as such, but only by the axon pathfinding algorithm. Moreover, this is a very simple algorithm that is embedded in the very process of brain growth, in the process of dividing its cells.
When brain cells divide, the execution of their genetic program changes from time to time according to a certain rule. This is how different areas of the brain are formed, consisting of groups of related neurons.
These areas are hierarchically nested within each other, so that a neuron of any “family” receives a certain place in space, depending on its ancestors. That is, the brain becomes the physical embodiment of the family tree of its cells.
Next, neurons begin to stretch with axons to “cousin” relatives of about the same generation. To find the “cousin” of its neuron, the axon simply reads the rule for changing the genetic program.
So the axon receives a sequence of neurons, next to which it needs to pass in order to find the right relative. Moreover, the physical location of each neuron is known in advance – it reflects its position in the genealogical tree of cells.
The axon begins to grow in the right direction, trying to find the desired neuron by molecular marks. As soon as the labels converge as much as possible, and further movement no longer improves the result, the axon “understands” that it has reached the desired neuron, discards the current label, reads the next one from the genetic rule and grows in a new direction.
Going through the labels in this way, the axon stretches from neuron to neuron. When it encounters a molecular label that is as similar as possible to the label of its neuron, this means that the “cousin” has been found and there is no need to grow further. The axon is connected to this neuron by synapse contacts.
To put it even more simply, the axon paves the path of growth in the same way that a person paves the way to a certain address in the city. But buildings and neighborhoods in a city can be randomly scattered, while clusters of neurons in the brain are always lined up in genealogical order.
This means that the axon does not even need a map, like a person, it is enough to move sequentially from point to point.
The researchers were able to model the development of the mouse brain in this way from the embryonic period (age 11.5 days) to childhood at 56 days (about a six-year-old child on a human scale).
At the same time, some of the simulated neurons connected with their nearest neighbors, while the other part grew long axons in order to reach their distant relatives. As it happens in the real brain.
Scientists acknowledge that the new model cannot yet fully describe the development of such a complex brain as the human one.
But they think their work does a pretty good job of explaining how an organ made up of neural networks and capable of learning can even form. A set of simple rules may suffice for the appearance of the most complex living structure.
Contact us: [email protected]