Imagine taking a map of a city, shredding it into 7,400 pieces, and then putting it back together. You have to align the landmarks, fix any tear or rip you’ve created in the process, and make sense of any shred you might have lost.
Now imagine doing the same to a preserved human brain. That’s what scientists at the Institute of Neuroscience and Medicine in Jülich, Germany and the Montreal Neurological Institute in Canada did. It was a monumental undertaking, and one that will undoubtedly contribute to President Obama’s recently announced BRAIN Initiative, which stands for Brain Research through Advancing Innovative Neurotechnologies. But when compared with the lofty goals of BRAIN, BigBrain, as this project is called, is just a small step down a very long and uncertain road.
BigBrain isn’t a part of the BRAIN Initiative, but it is precisely the kind of research that is needed to advance the multibillion-dollar, multiyear federal initiative toward its goal of mapping all the neurons in the human brain. The ambitious BRAIN Initiative is often compared with the Human Genome Project, a research program that had similarly lofty goals and a long timeline. It took the Human Genome Project 13 years to sequence our entire genome, and that’s after the project was officially started. The BRAIN Initiative is still in the planning phases, with timelines and specific goals still being formulated. If anything, the work on BigBrain, detailed in this week’s issue of the journal Science, puts into perspective the enormity of the BRAIN Initiative.
As the name implies, BigBrain wasn’t a small side-project. Before it, most studies of its kind used the mouse brain, which is 7,500 times smaller than the human brain. Completing BigBrain took a team of scientists 10 years. They started by obtaining the brain of a healthy, 65-year-old woman who died of non-neurological reasons. Before they cut into it, they put the brain through an MRI scanner. That provided them with a reference—like the lid of a puzzle box—that they could use later when reassembling the brain. The team then sliced it into 20-micron sections. (By comparison, a human hair is about 70 microns thick, and a piece of paper is about 100 microns.) Alan Evans, one of the study’s authors, likened them to Saran wrap, except they didn’t come in a single roll but over 7,400 individual sections.
Next, they dyed each section to distinguish different cells, and then scanned the sections into the computer, a task that took 1,000 hours to complete. Once digitized, researchers repaired rips and distortions that occurred during slicing and scanning. Finally, they put the entire brain back together, slice by slice, using the MRI as a guide.
Zooming In: A Map of Houses
Prior to this study, MRI provided the most detailed 3D peek into a human brain. If you think of the brain as a map of a country, the resolution of MRI—about 1 millimeter—would make towns visible, but nothing smaller than that would be. BigBrain, on the other hand, “does 50 times better in each dimension than the typical 1-millimeter resolution of MRI,” says Katrin Amunts, a neuroscientist at the Institute of Neuroscience and Medicine in Jülich, Germany, and lead author of the paper. Specifically, BigBrain’s 20-micron resolution is fine enough to pick out individuals of certain types of cells, but not all; the smallest neurons in the brain are only about 10 microns across. Still, if this were a map, the level of detail provided by BigBrain greatly exceeds MRI, allowing us to see not just towns, but the houses within them.
“It’s an extension of the classic work that’s been done literally since the turn of 20th century to try to provide a more accurate description of the various areas of the brain and the morphology of the brain,” says Donald Stein, a neuroscientist at Emory University who was not involved with the work. Early in the 20th century, neuroscientists painstakingly surveyed the brain and documented what the cellular populations looked like, the various shapes, sizes, and groupings of different cells in different parts of the brain—a cul-de-sac of small houses here, large farmhouses scattered across a countryside over there. Certain regions of the brain, known as Brodmann’s areas, were later discovered to have specific functions, such as language, vision, or hearing. Today, BigBrain gives us a clearer picture of these areas.
But it’s still just one person’s brain, says Joshua Sanes, a neuroscientist at Harvard University and a member of the BRAIN Initiative advisory committee. He doubts that we will learn a whole lot by just looking at one sample. But he’s optimistic, explaining that BigBrain can be used as a reference to help ensure consistency across different studies and subjects. For example, a study comparing brains from healthy people with brains from Alzheimer’s patients would need to compare them to a standard reference brain in order to correct for the typical variability between subjects. Sanes hopes BigBrain will fulfill that role.
It can also serve as a scaffold, supporting data from other studies, just as the frame of a house supports siding and shingles. Currently, functional MRI can tell us which general region is involved in a function, like reading, but it doesn’t give us much more than a hazy outline. Other types of studies may use BigBrain to mark the location of different neurotransmitter receptors or patterns of gene expression. Allan Jones, the CEO of the Allen Brain Institute in Seattle and not an author on the paper, says BigBrain will be able to “anchor all of the great science being done in imaging to a much higher resolution atlas and framework.”
Not So Fast: Life Without Streets
Most scientists I interviewed agree this is a great start, but some still question whether BigBrain could live up to the hype surrounding its introduction. Stein likens BigBrain to “the high resolution screen on the latest iPhone. Is it really that much better than the last iteration that came out six months ago?”
Perhaps the biggest pieces missing from BigBrain are the connections between neurons and the different regions of the brain. Right now, BigBrain is like a map of a city with no streets, just buildings. While that sounds like a glaring oversight, Sanes says it’s a sensible approach; if researchers had tried to trace these physical connections by staining them, the brain would have emerged an inscrutable, tangled mess. To make the next step—understanding how neurons communicate and connect with each other—we’ll need additional studies that use different technologies. But even here, BigBrain can help. Michael Eisen, a biologist at the University of California at Berkley, suggests that we can begin to look at traffic patterns in the brain by overlaying other types of imaging onto the BigBrain atlas.
“There are a lot of other equally valid approaches to understanding how the brain might work and [BigBrain] is just another one in the armamentarium that we should all be able to use and pick and choose from.” Stein says. Eisen agrees that BigBrain is one more tool in the chest. “Collectively, all these things are going to be important in moving the field forward, but any individual one of these things [isn’t] going to revolutionize the world.”
Zooming Out: The Bigger Picture
BigBrain is an early but important contributor to the BRAIN Initiative, Sanes says. This new paper, “reported one step on a very long path toward understanding the brain,” he says. “I think of it largely in terms of long-term goals. For example, this BRAIN Initiative.”
What the BRAIN Initiative hopes to accomplish is far beyond what neuroscience can do today. Currently, scientists can count the neurons of simple animals like worms and map some detailed circuits in mouse brains. Sanes suggests the early stages of the BRAIN Initiative will be similar to the Human Genome Project. When that research program began in 1990, genomes of smaller organisms like bacteria and worms had been sequenced, but the human genome remained a daunting and expensive task. The early stages of the genome project involved developing new technology that would ultimately make the rest of the undertaking much cheaper and easier, Sanes says.
Neuroscience is in much the same position today. Many technologies that will be needed to accomplish the BRAIN Initiative’s goals, like those behind BigBrain, are in their infancy. When the BRAIN Initiative begins next year, BigBrain will serve as a reminder of the difficulty of each step toward the ultimate goal.
Eisen, though, is optimistic. He predicts that we’ll see similar advances in the coming months. “This is very much a field in a dynamic moment in time,” he says. “This is what it looks like when a field is really humming.”