Imagine a map with every star in the galaxy. The map is so detailed that it shows what each star looks like, what it consists of, with which other star is connected by the great laws of the physics of space. Although we do not yet have such an astronomical map of the heavens, thanks to a monumental study published last week in Neuron, we have a similar brain map.
If every neuron were a galaxy, synapses are small structures scattered in dots along serpentine neurons - these are stars. A team of scientists from the University of Edinburgh in the UK has built the first detailed map of each synapse in the mouse brain.
Brain map: the key to the puzzle of thinking
Using genetically modified mice, the scientists literally made each synapse light up under fluorescent light all over the brain, like stars in the sky. And, just as stars differ from one another, scientists have found that synapses are very diverse, but patterns are observed that can support memory and thinking.
Detailed maps showed the fundamental law of brain activity. Using machine learning, the team has categorized the roughly one billion synapses throughout the brain into 37 types of synapses. Here's the point: when sets of neurons receive electrical information, for example, choosing between different solutions to a problem, the unique subtypes of synapses scattered among the different neurons unanimously sparkle with activity.
In other words, synapses are of different types. And each type can control a thought, decision, or memory.
Promotional video:
Unsurprisingly, neuroscientists have reacted very positively to the work.
"Wow," commented Ben Sanders of the University of Minnesota.
It is "an amazing article cataloging the diversity and distribution of synapse subtypes throughout the mouse brain," writes neurogeneticist Kevin Mitchell. This "underscores the fact that synapses are key computing elements of the nervous system."
Connect the connectome
Scientists' interest in creating a "synaptome" - the first complete catalog of synapses in the mouse brain - stemmed from a much larger project: the connectome.
In short, a connectome is all the neural connections within you. As Dr. Sebastian Sjung says, the connectome is the biological basis of who you are - your memories, your personality, your thoughts and reasoning. Catch a connectome, and one day scientists will be able to reconstruct you by emulating a whole brain.
And yet, the connectome only describes how neurons functionally talk to each other. Where in the brain is it physically encoded.
This is where synapses come into play. Neuroscientists have long known that synapses transmit information between neurons using chemicals and electricity. There have also been hints that synapses differ greatly depending on the proteins they contain, but this distinction has usually been ignored. Until recently, most scientists believed that real calculations took place in the neural body - the bulbous part of the neuron from which the branches emerge.
Until now, there was no way to look at the morphology and function of synapses throughout the brain, the authors explain. We have usually focused on mapping these important connection points in small areas.
And if so, future brain emulators will finally find a foothold.
Synaptome map
To build a mouse synaptome, the authors devised a plan called SYNMAP. They started with genetically modified mice, whose synapses glowed with different colors. Each synapse is densely packed with different proteins, among which PSD-95 and SAP102 are the most famous gentlemen. Don't be intimidated by the names. The authors added glowing proteins to them, which acted as flashlights, illuminating each synapse in the brain.
In general, scientists first altered the biology of the mouse by making its synapses glow under fluorescent light.
They then painstakingly sliced the brain into pieces, used a microscope to take pictures of synapses in different regions of the brain, and put the pictures together.
The synaptic image reminds the untrained eye of a tightly packed star map - like the one that Hubble recently filmed. The categorization of each synapse goes beyond the ability (and time) of any person, so scientists have used new classification methods using machine learning and developed an algorithm that parses this data - more than 10 terabytes - automatically, without the need for supervision.
Physical connectoma
At first, scientists were struck by the "expressive schemes" of the shining synapses. One labeled protein - PSD-95 - seemed to hang out in the more distant parts of the brain, where higher cognitive functions take place. Although the regions overlap, another glowing protein has preferred more of the inner brain regions.
Upon closer inspection, it turned out that the two glowing proteins represent different sets of synapses, the authors explained. Each area of the brain has a characteristic "synaptic signature." Like fingerprints, which differ in shape and size, different regions of the brain seem to contain synapses that differ in protein composition, size and number.
Using a specially designed machine learning algorithm, the scientists classified the synapses into 37 subtypes. Remarkably, the brain regions associated with higher reasoning and thinking abilities also contained the most diverse population of synapses, while the "reptilian brain regions" were more uniform in synaptic content.
To see if the diversity of synapses helps in information processing, the scientists used computer simulations to show how synapses respond to normal electrical circuits in the hippocampus, an area of the brain important for learning and memory. The hippocampus is one of the areas that show striking diversity in synapse subtypes.
Importantly, each type of electrical information processing is transmitted to a unique synaptome map - change the input, the synaptome will change.
This suggests that the brain can process multiple electrical information using one area of the brain because different synaptoses are involved.
Scientists found similar results when they recorded electrical circuits of the brains of mice trying to choose between three reward options. Various synaptomes were fired when the choice was right or wrong. Like a map of inner thinking, synaptomes painted a vivid picture of what the mouse was thinking when making a choice.
Each behavior activates a separate synaptome. Each synaptome is a unique cast of the thought process.
Reprogramming the synaptome
Like computer code, the synaptome appears to underlie a computational outcome - a decision or a thought. What if I change this code?
Psychiatric illnesses often have genetic causes that affect proteins in the synapse. Using mice that showed symptoms similar to schizophrenia or autism, the scientists mapped their synapses - and found dramatic changes in how different subtypes of synapses in the brain are structured and connected.
For example, in response to normal electrical circuits in the brain, some synaptic maps were weak, while others became abnormally strong in mutant mice. Mutations can change the synaptome and potentially lead to psychiatric disorders. That is, some psychiatric illnesses "reprogram" the synaptome. Stronger or simply new synaptic maps can be the reason schizophrenic patients experience delusions and hallucinations.
So you are your synaptome?
Maybe. The essence of you - memories, thoughts - seems to be imprinted in how different synapses are activated in response to input. Like a fingerprint, the synaptome could be read to decipher what you are thinking. However, this research is just the beginning. Neuroscientists have yet to analyze the complex connections between synapses and you.
Ilya Khel