Connectomics: Unraveling the Secrets of the Brain
The Exciting Science Behind Generating A Map of the Brain
Humanity has come a long way since we first evolved from our ape ancestors. It started slow (extremely slow), with figuring out how to light a fire and make tools using stones, but quickly picked up to the point where, about 200,000 years later, we can manipulate life at the smallest orders of molecules, launch people into space, and even manipulate atoms to meet our desires.
But, in all this time, if there’s one thing that we continue to barely understand, it’s, ironically, the part of our body we use the most to understand everything about the world: our brain. Up to now, this extreme gap in our knowledge of the brain was largely due to the lack of the right tools and knowledge for most of our history. The human brain is composed of around 100 billion cells, with more than 1 quadrillion connections between them, with each connection sending 1–1000 signals every second. With such orders of magnitude, mapping every connection and completely understanding the brain is an unimaginable task that understandably could never be accomplished simply by looking at neurons under normal, school-grade microscopes.
Of all the parts of the body that humans have studied, there is none more complicated and mysterious than the brain. As biologists move into different terrains, diseases of the nervous system still remain embarrassingly poorly understood… and the physical and functional architecture of the brain is understood much less than any other organ. — Jeff Lichtman, Center for Brain Science, Harvard
But, with the rapid innovation and development of computing technologies, we may finally have the necessary tools to finally start looking at the huge amounts of data just waiting to be analyzed in the brain. And researchers around the globe are now committed to using these technologies to finally crack the brain, establishing the field of connectomics.
The Brain Is Mind-Blowing (Pun Totally Intended)
Quick Crash Course Neuroscience
Before delving into the science of connectomics, we need to first understand the basic structure and function of the brain, and why we need connectomics in the first place.
Generalized, the brain consists of two main types of nerve cells: neurons and glia. Neurons, making up just 10% of the brain, are the cells responsible for the actual function of the brain, by sensing changes in the environment, communicating these changes to other neurons, and commanding the body’s responses to sensations. Glial cells, making up about 80–90% of the brain, contribute to the brain’s functions mainly by insulating, supporting, and nourishing the neurons, making sure that they can function properly. But each of these large categories themselves consist of numerous subcategories. In fact, even today, scientists continue to dedicate their careers to dissecting the brain and finding new types of neurons and glia.
Adding onto the number of cells that need to be analyzed are the millions of connections between these neurons. Everywhere else in the body, the function of the organ is largely represented by the individual cellular units themselves. If one cell in the heart is lost, the function won’t be impaired, since there are thousands of other cells performing the same function. But in the brain, the function is represented not by the work of individual neurons, but by the connections allowing multiple neurons to work in harmony. Each neuron has two processes extending from its cell body: dendrites and axons. Dendrites receive signals from other neurons, conducting them to the cell body. Axons, on the other hand, relay messages to neighbouring neurons. Where an axon from one neuron meets a dendrite from another, the connection between the two is referred to as the synapse, where chemical transmitters released from the axon terminal are detected by receptors in the dendrite, transferring information between the neurons.
Because of this unique structure of neurons, the proper functioning of the brain requires all the neurons as a whole to work well together. So understanding the brain requires not only understanding how all the different types of nerve cells work but also how they are all connected to one another and how specific connections give rise to specific functions in the brain. And when there are hundreds of trillions of synapses, as estimated in the human brain, understanding all those connections becomes nearly impossible.
To understand the function of the brain, you really have to understand how they [nerve cells] are connected.
What Is Connectomics?
Despite the complexity, researchers like Dr. Jeff Lichtman are attempting to breakdown this structural complexity of the brain by using computer-assisted brain imaging to map the structures and connections in the entire brain.
There are two main imaging methods currently being used for this, relying on light microscopy and electron microscopy, respectively.
Light Microscopy, & Brainbow
Light microscopy uses various techniques involving (as you might guess) light to reveal cellular structures too small for the naked eye. If you have ever been in a biology or science class, you have most likely used light microscopes to study microscopic cells. But traditional light microscopes found in school science labs lack the specificity and focal capacities required to truly decipher the structure of the nervous tissues. Though these types of microscopes can identify single cells via staining, they make it difficult to identify and follow individual processes extending from one neuron to the next. In order to make any progress using light microscopy, scientists need to use clever techniques that allow them to extract more information about the structure of the brain.
One such innovation came from the Lichtman lab (leaders in the connectomics field), which they called Brainbow. The technique used fluorescent proteins (XFPs) in order to label every single cell in the brain with a different colour, hoping to reveal all of the different neurons and their separate processes. Essentially, they introduced several different genes for red, green, and blue (RGB) XFPs into neurons, so that each neuron would have different levels of each of the three proteins. This method, which took inspiration from screens like the one you’re looking at right now, allowed the creation of virtually every colour possible and separate labelling of each and every neuron.
Though the technique produced extremely beautiful images of the brain, scientists were still unable to distinguish and track individual processes in the dense clusters that can be seen extending from the cell bodies. If researchers really wanted to decipher the structure of the brain, they needed a way to zoom in even further. And electron microscopy was just the way to do that.
Electron microscopy and ATLUM
Electron microscopy (EM) is a technique for obtaining high-resolution images using electrons, which have very short wavelengths, as the source of the illuminating radiation. This method allows scientists to image much smaller structures with much higher detail and resolution.
The problem with EM however is the fact that it can only image an extremely small tissue section at a time, requiring researchers to slice up the brain into thousands upon thousands of tiny slides. But in this case, scientists are not simply looking for the structure of specific components, but actually at how the neurons are spatially organized and connected to each other. And to be able to decipher this kind of information, the slices and images must be coordinated in a way that allows for a program to be able to perfectly piece together each piece to maintain its order and organization in the brain.
Once again, the Lichtman lab has come up with a solution to this problem, by developing the automatic tape-collecting lathe ultramicrotome (ATLUM), which slices flesh samples into extremely thin slices to allow scanning under an electron microscope. These slices are then automatically collected on a long carbon-coated tape, for later analysis under an SEM. The big advantage of this machine is the ability to automate the slicing process to create ultrathin sections of volumes as large as tens of cubic millimetres quickly and reliably (this might seem like a very small amount, but these volumes can span entire multi-region neuronal circuits). Furthermore, the order of the slices, and thus the organization of neurons, is conserved, allowing for the scans to be pieced together to form a proper image of the brain.
Analyzing the Images
Now that we have the images from EM, how can we make sense of it? Well, this step was actually quite cumbersome earlier in the connectome project, and still remains difficult and complex. Early on, scientists and researchers (usually student researchers in the lab) would manually analyze tissue sections and trace each and every neural connection from image to image. Though highlighting and tracing these neural connections from image to image is still needed, it now involves more computer-assisted processes that speed up the process immensely. On top of that, researchers like Sebastian Seung have also made some of the data and images open-source so that the common Joe can go in and start identifying neuronal connections in EM scans in these images. You can try this for yourself at Eyewire (which I totally did not spend a couple of hours playing around with 🙄).
Despite this, the amount of data being produced by researchers grows at a rate much faster than it can be analyzed. According to Dr. Lichtman, the dataset representing the neural wiring of just the mouse brain, which has 1000-fold fewer cells than the human brain, will be on the scale of an exabyte, or one billion gigabytes. Contrast this to the human genome, the entirety of which can be represented in just 1.5 gigabytes. So it’s safe to say that it’s unlikely that a complete connectome of the human brain will be developed any time soon. But, despite its challenges, the project is still extremely crucial, since fully understanding the structure of our brain can allow us to diagnose and treat neurological disorders with much higher precision and accuracy than possible with today’s understanding of the brain.
So until the day when scientists finally announce that they have mapped the entirety of the human brain, we’ll just have to stick with using MRIs and other much simpler techniques to advance our understanding of the brain.
Summary
- The human connectome project aims to decipher and map all the connections between neurons in the human brain.
- The research currently makes use of light microscopy and electron microscopy techniques, with different innovations for each technique to increase the amount of information that can be extracted from these images.
- The amount of data predicted to be produced by the connectome project is on an unfathomable scale, representing the complexity of the brain and the difficulty of deciphering its structure.
- Having a full connectome can allow us to develop more effective and accurate diagnostics and treatments for a host of neurological disorders.
On a Personal Note
This is my first article on Connectomics, and I really appreciate you taking the time to read it. If you liked it, please clap for the article!
To stay updated with more of my articles (about different areas of technology and medicine, including lots of neuroscientific research), follow this Medium account, check out my personal website to find out a bit more about me, and follow me on my LinkedIn and Twitter to stay updated.
That’s it for me! See you next time!
