IT’S A LONG, hard road to understanding the human brain, and one of the first milestones in that journey is building a … database.
In the past few years, neuroscientists have embarked on several ambitious projects to make sense of the tangle of neurons that makes the human experience human, and an experience. In the UK, Henry Markram—the Helen Cho to Elon Musk’s Tony Stark—is leading the Human Brain Project, a $1.3 billion plan to build a computer model of the brain. In the US, the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative hopes to, in its own nebulous way, map the dynamic activity of the noggin’s 86 billion neurons.
Now, the Allen Institute for Brain Science, a key player in the BRAIN Initiative, has launched a database of neuronal cell types that serves as a first step toward a complete understanding of the brain. It’s the first milestone in the Institute’s 10-year MindScope plan, which aims to nail down how the visual system of a mouse works, starting by developing a functional taxonomy of all the different types of neurons in the brain.
“The big plan is to try to understand how the brain works,” says Lydia Ng, director of technology for the database. “Cell types are one of the building blocks of the brain, and by making a big model of how they’re put together, we can understand all the activity that goes into perceiving something and creating an action based on that perception.”
The Allen Cell Types Database, on its surface, doesn’t look like much. The first release includes information on just 240 neurons out of hundreds of thousands in the mouse visual cortex, with a focus on the electrophysiology of those individual cells: the electrical pulses that tell a neuron to fire, initiating a pattern of neural activation that results in perception and action. But understanding those single cells well enough to put them into larger categories will be crucial to understanding the brain as a whole—much like the periodic table was necessary to establish basic chemical principles.
Though researchers have come a long way in studying the brain, most of the information they have is big-picture, in the form of functional scans that show activity in brain areas, or small-scale, like the expression of neurotransmitters and their receptors in individual neurons. But the connection between those two scales—how billions of neurons firing together results in patterns of activation and behavior—is still unclear. Neuroscientists don’t even have a clear idea of just how many different cell types exist, which is crucial to understanding how they work together. “There was a lot of fundamental information that was missing,” CEO Allan Jones says in a video about the database project. “So when we got started, we focused on what we call a reductionist approach, really trying to understand the parts.”
When it’s complete, the database will be the first in the world to collect information from individual cells along four basic but crucial variables: cell shape, gene expression, position in the brain, and electrical activity. So far, the Institute has tracked three of those variables, taking high-resolution images of dozens of electrically-stimulated neurons with a light microscope, while carefully noting their position in the mouse’s cortex. “The important early findings are that there are indeed a finite number of classes,” says Jones. “We can logically bend them into classes of cells.”
Next up, the Institute will accumulate gene expression data in individual cells by sequencing their RNA, and the overlap of all four variables ultimately will result in the complete cell type taxonomy. That classification system will help anatomists, physicists, and neuroscientists direct their study of neurons more efficiently and build more accurate models of cortical function. But it’s important to point out that the database isn’t merely important for its contents. How those contents were measured and aggregated also is crucial to the future of these big-picture brain mapping initiatives.
To create a unified model of the brain, neuroscientists must collect millions of individual data points from neurons in the brain. To start, they take electrical readings from living neurons by stabbing them with tiny, micron-wide pipettes. Those pipettes deliver current to the cells—enough to get them to fire—and record the cell’s electrical output. But there are many ways to set up those electrical readings, and to understand the neural system as a whole, neuroscientists need to use the same technique every time to make sure that the electrical traces can be compared from neuron to neuron.
The Allen Institute, in collaboration with other major neuroscience hubs—Caltech, NYU School of Medicine, the Howard Hughes Medical Institute, and UC Berkeley—has made sure to use the same electrical tracing technique on all of the neurons studied so far (they call it “Neurodata without Borders”). And while the data for this first set of mouse neurons was primarily generated at the Institute, those shared techniques will make future work more applicable to the BRAIN Initiative’s larger goals. “In future releases, we’ll be working with other people to get data from other areas of the brain,” says Ng. “The idea is that if everyone does things in a very standard way, we’ll be able to incorporate that data seamlessly in one place.”
That will become increasingly important as the Institute continues mapping not just mouse neurons, but human ones. It’s easy to target specific regions in the mouse brain, getting electrical readings from neurons in a particular part of the visual cortex. It’s not so easy to get location-specific neurons from humans. “These cells actually come from patients—people who have having neurosurgery for epilepsy, or the removal of tumors,” says Ng. For a surgeon to get to the part of the brain that needs work, they must remove a certain amount of normal tissue that’s in the way, and it’s that tissue that neuroscientists are able to study.
Because they don’t get to choose exactly where in the brain that tissue comes from, scientists at Allen and other research institutes will have to be extra careful that their protocols for identifying the cells—by location, gene expression, electrical activity, and shape—are perfectly aligned, so none of those precious cells are wasted. All together, the discarded remnants of those human brains may be enough to reconstruct one from scratch.