by Olivia Ho-Shing
figures by Tito Adhikary
Imagine yourself as a self-sustaining city. Everything you think and all the tasks you carry out are driven and coordinated by a massively complex power network. This is the brain. It transmits thousands of electrical signals along circuits of intricate power lines within milliseconds to execute your behaviors. How can you begin to understand these network circuits that shape those behaviors?
The brain is in many ways an uncharted terrain, covered with a dense web of neural circuitry. From early work in neuroscience, we understand that the dense web of the nervous system is made up of discrete, individual cells called neurons, which form circuits via specialized connections. Neurons are the power lines that stretch across the terrain of the mind. To truly understand how the brain works, we need to understand the structure and activity of the individual neurons. A good place to start is to map out how the network is organized and connected – this is the challenge of mapping the brain.
We understand the world most clearly through seeing it. Visualization is key for studying the brain because sight is arguably our strongest human sense. With this in mind, the Brain Activity Mapping (BAM) project, proposed in 2013, is an ongoing collaborative effort between scientists to paint real pictures of brain circuitry. BAM aims to build neuroscience tools capable of measuring large sets of neural activity, as well as computationally analyzing and modeling these neural circuits, and testing the resultant models. This article highlights only some of the revolutionary advances in imaging and molecular techniques that enable researchers to visualize the brain at the nanoscale – the scale at which individual neurons develop, connect, and fire – in order to form a reference map to help us navigate through the brain.
Until recently, traditional approaches could only track the electrical activity of few neurons at a time, and functional imaging studies like MRIs done on humans could only survey the activity of large general brain regions. If we understand the brain as a grid of power lines, these traditional approaches can be likened to testing each power line one by one, or tracking power usage by state— methods that provide insight into activity on a very small or a very large scale, while not presenting a full picture of the complete neural circuitry. While we have insight into how the network as a whole is working, we are afforded only limited perspectives on how well a local circuit may or may not be working.
However, if we could visualize every power line or neural connection, we could learn which ones worked together, what they do with signals they receive, and even get hints at where they are prone to stop working. New research tools and techniques to image the brain are particularly groundbreaking because instead of tracking electrical output, we now have a number of ways to directly view the structure and activity of neuronal circuits.
The first step in making a map of the network is basic, but incredibly important: we must chart out the locations of all the connections. In the late 1800s, neuroscientist Santiago Ramón y Cajal helped to definitively establish the understanding of how neurons are organized by improving a method to stain sparse numbers of neurons in brain tissue. He used silver nitrate and potassium dichromate to fill the neuron with a dark stain, then drew detailed illustrations of the neurons he saw through his microscope. By staining only a few neurons in brain tissue, Ramon y Cajal was able to see the intricate structure of different neuronal cell types. His illustrations are seminal in neuroscience because they provide the first views of the microscopic structure of the brain.
However, Ramon y Cajal’s pictures only illustrated a small subset of all the neurons that make the brain function. The problem in getting a complete snapshot is that the brain is incredibly jam-packed with different kinds of cells. In order to chart the course of each neuron, researchers needed a strategy to see them all, and still distinguish one from the other. To tackle this problem, leading investigators genetically introduced fluorescent proteins into every neuron. Along with the normal proteins the neuron makes, each neuron then expresses a random combination of the primary colors red, green and blue. These colors combine to paint each neuron a color unique from its neighbors. Using this method, called Brainbow, a single neuron and its connections can be tracked, much like a distinct color band in a rainbow. Imaging sections of the Brainbow mouse brain results in beautiful never-before-seen images that visualize the scope and complexity of neuronal connections.
Although Brainbow can only be applied to studying brains of model organisms like mice, the clear insight this method gives researchers into brain structure is key in advancing the field of neuroscience. Because Brainbow permanently gives a neuron and all of its descendants a unique color, this technique also help researchers study how neurons develop and change with age or injury. For example, researchers used Brainbow to track the regeneration of axons, the neuron’s long projection that sends the signal to the next neuron. They found that after nerve damage, debris stuck around the axon slows down and can obstruct neuron regeneration. By charting out the full picture of neurons, we have insight into how the system breaks down, and perhaps what we can do to combat that in the future with medicine.
Gaining CLARITY in Tissues
Now that we have fluorescent techniques like Brainbow to make sense of the crowdedness of the neurons, there is another problem in completing our map. Unlike a two-dimensional map of power lines across a terrain, mapping the brain requires considering the fact that it exists in three dimensions. Neurons organize themselves and connect to each other across long ranges, so to trace them and find out what they do, we need to be able to see them through a volume of space.
Often, 3D imaging is constructed by stacking multiple two-dimensional images on top of each other to reconstruct a three-dimensional volume of tissue. This is how many Brainbow images are pieced together, but this method becomes extremely difficult as the number of pieces to reconstruct increases. Recent developments in optic techniques have introduced new advanced microscopes, like two-photon and lightsheet illuminations, which allow researchers to image deeper into the whole brain instead of just brain slices.
A second underlying challenge in imaging neurons in the brain is the inevitable difficulty in seeing through the mesh of tissues supporting the neurons. In fact, most biological tissues have an inherently milky appearance that obscures the sight of the cellular structures when attempting to image deeper than an object’s the surface. Even with strategies like Brainbow that colorfully label neuronal connections, many neuronal cell types have multiple fine connections that are easily lost. However, being able to trace these connections is integral to completing a map and understanding where each circuit leads.
To address this problem, new chemical and molecular approaches have gained a great deal of traction in making samples of intact brains physically transparent. CLARITY, iDISCO, and Scale are a few of the new techniques introduced to clear brain tissue, pioneered in mouse brain. The process of making brain tissue more transparent is similar to making milk clearer—it requires removing fats and proteins, which is why skim milk appears more transparent than whole milk. After proteins and fats are removed, the whole brain is then embedded in a gelatin-based substance to help maintain the brain structure for imaging.
In a cleared brain, individual neuronal circuits can then be imaged without slicing it apart. Clearing away the background can also illuminate other structural aspects of neuronal diseases. For example, researchers at the RIKEN Brain Science Institute in Japan used Scale to visualize amyloid plaques, which develop around the blood vessels and contribute to neuron degeneration in mouse models of Alzheimer’s disease.
The ability to visualize all of the elements in a circuit allows researchers to trace out these connections to learn which neurons are communicating with each other. By working to detangle these lines of communication, we will be able to better identify degenerative conditions like dementias, wherein these connections tend to be lost.
Imaging Small Brains
Lastly, while having a clear picture of neurons in brain tissue is beneficial for characterizing connections and structural dysfunctions in the brain, to understand how a brain network controls how we eat or even read this article, our map needs data on which neurons are actively communicating. Tracking the patterns of activity across neural connections provides an understanding of the kinds of behavior a circuit controls.
The challenge, however, in achieving this in the brain is finding a reliable readout of activity in large populations of neurons. Given the high levels of activity across the network, researchers need a way to record clear immediate reports of activity from individual neurons. The traditional method to track a neuron’s activity is to record its electrical currents; while this is a very immediate report, it cannot be tracked for more than a few neurons at a time. Now, researchers can instead track activity by chemical changes that occur in the neuron when it is activated. Activated neurons increase their calcium concentration, which immediately triggers neurotransmitter release, transmitting the signal to the next neuron. Researchers have developed a protein that responds to an influx in calcium by changing shape and fluorescing. This new tool gives neuroscientists a visible readout of neuronal activity allowing them to image large populations of neurons at the time they are active in the model brain.
While visualization of individual neuronal activity is still difficult to accomplish in a brain the size of a mouse’s, because of the sheer volume of neurons and the consistent and rapid nature of network activity, there are smaller brains researchers can visualize more completely to gain insight into common network behaviors. Current model organisms where imaging can encompass the entire brain are the nematode (C. elegans), the fruit fly (Drosophila melanogaster), and the zebrafish (Danio rerio). The largest of these organisms is still smaller than a finger, so what can their neural processing teach us about ours as humans? Despite their size, all of these animals, even the nematode, which has 302 neurons in total, can perform basic behaviors like movement as well as complex behaviors like learning and aggression. Neuroscientists can test and manipulate these simpler brain circuits to learn principles that can be generalized to human behavior. For example, mapping neural activity in fruit flies given anesthetics can inform us about principles of basic drug tolerance in our own brains.
So what is the ultimate purpose of creating these detailed maps of the brain? A complete map enables us to understand the terrain and develop ways to improve how well our personal power grids function. Development of these new tools and strategies is empowering neuroscientists to gain more sophisticated insight into the intricate functions and organization of the most complex network we know. The insights we gain from research like this will directly improve our approaches in medicine, as well as our understanding of brain development and neurodegenerative diseases. Although it is a massive daunting task to decipher the brain, each innovation and new insight can improve how we approach mental health and developmental learning.
Olivia Ho-Shing is a PhD student in the department of Molecular and Cellular Biology at Harvard.
This article is part of the April 2016 Special Edition on Neurotechnology.
On BAM and the Challenges of Mapping: http://www.sciencedirect.com/science/article/pii/S0896627312005181
Brainbow Picture Gallery: http://www.cell.com/pictureshow/brainbow
On Interpreting Activity Maps into Insights: http://www.cell.com/neuron/fulltext/S0896-6273%2814%2900796-X