By: Greg Miller
The Obama administration wants to make a huge investment in mapping the human brain, according to The New York Times. How can they get the most bang for their buck? We have details on three future technologies that are being eyed by the scientists behind the bold proposal.
The U.S. already has one big brain-mapping effort under way, the Human Connectome Project, which aims to map the connections between regions of the human brain. The new project would go beyond this static depiction and map the activity of individual neurons in real time.
“All the really interesting features of the brain — language, perception, cognition, the mind — emerge from collections of neurons interacting with each other in ways we don’t understand,” said neuroscientist John Donoghue of Brown University, one of the architects of the proposed project. It’s those interactions, the electrochemical blips coursing through networks of interconnected neurons, that the new Brain Activity Map project aims to capture.
The Connectome project focuses mostly on static images of the brain. Although it does include some measures of brain activity, the fMRI scans it will use provide a view that’s something like that of a city seen from an airplane window. What the scientists behind the proposed Brain Activity Map want instead are detailed street maps with real-time traffic info. Ideally, they want to record every blip of every neuron in a network of thousands, or even millions.
The scientists hope they’ll get as much as $3 billion over the next decade to build a new set of dream tools for studying the brain. At first these tools could only be used with lab animals, but ultimately they could help unravel how the human brain works and what goes wrong in disorders like epilepsy and Alzheimer’s disease. Here are three ideas they’ve discussed, all in various stages of development.
“Sure, they sound far-fetched,” Donoghue said. “But we’re on the cusp of being able to do them.”
For decades, the workhorse method for recording the activity of individual neurons has been hair-thin metal electrodes. Not only are they invasive — like sticking a toothpick into a bowl of Jello — but they only record from one neuron at a time. More recently, scientists have built grids with dozens of electrodes. Donoghue’s team, for example, has shown that signals from just 100 neurons or so are enough to allow a paralyzed person to operate a robotic arm.
But even that may start to look crude by comparison. Harvard physicist Hongkun Park is one of several scientists trying to pack hundreds of thousands of nanowire electrodes into flexible sheets that conform to the surface of the brain and can eavesdrop on neurons with minimal tissue damage. “I didn’t know we could make such things,” says Donoghue, who saw Park talk about his work at an early planning meeting for the brain-mapping project.
Optical methods have already revolutionized neuroscience in the past decade, providing new ways to record and manipulate the activity of neurons with light. But that’s just a start, says Rafael Yuste, a neuroscientist at Columbia University and one of the scientists behind the proposed mapping project. “We’ve just seen an appetizer of what optics can do in the brain,” Yuste said.
Imagine that your brain is a giant TV with 100 billion pixels and it’s playing a movie. “Each pixel is a neuron and that movie is your mind,” Yuste said. With a conventional electrode, a neuroscientist can watch one pixel/neuron at a time, and she’d have no clue what the movie is about. Today’s state-of-the-art optical recording methods can capture 100 neurons or more, but there’s a time lag that blurs the picture. To even begin to be able to follow the plot, researchers need to follow 100,000 neurons with no time delay, Yuste says. That will require better chemical sensors for visualizing electrical activity, as well as better optical hardware.
Another huge obstacle is peering deeper into the brain. “For hundreds of years people have designed microscopes to focus on a single plane,” Yuste said. “We have to redesign the microscope to focus on all the planes at the same time so we can see in 3-D.” That will require help from physicists and engineers, which are exactly the kind of collaborations Yuste and others hope the new mapping project will inspire. “There’s enough ideas in the hopper to make me confident that major progress will happen,” he said.
Synthetic biology sensors
DNA can store vast amounts of data in a tiny space. Harvard molecular technologist George Church famously used it to store his book, and now Church says DNA also could be used to record neural activity. He estimates that the 3 billion base pairs of the human genome have enough storage capacity to record everything a neuron does for a week, or at least every electrical spike, assuming it fires 100 times per second.
One strategy for doing this, which his team described last year in PLoS ONE, involves enzymes called DNA polymerases, whose job it is to make exact copies of a DNA sequence. But some of these enzymes have a tendency to mess up in the presence of positive ions, which happen to be exactly what floods into a neuron when it fires. The mistakes made by the polymerase create a record of the neuron’s firing that could be read out later in the sequence of the copied DNA.
It’s a long way from being practical, but Church is already thinking about how to get the DNA recorders into the brain (smuggle them inside synthetic immune cells, perhaps) and how to recover the ticker-tape record of neural firing once it’s made (possibly reading it out with still-to-be invented optical methods that could even identify which neuron it came from).