Building Brains


The brain is no flat slab of circuitry. If you want to make one—without doing it the old-fashioned way—you'll need a technique for building in all three dimensions.

By and large, the vast majority of tissue growing in today's labs has been done in only two dimensions. Where a third dimension has been attempted, the structures have been limited and the costs prohibitive (stereolithography machines, for instance, cost tens of thousands of dollars). But now researchers at MIT and Harvard Medical School have developed a new way to build three-dimensional brain tissues in any conceivable shape—and on the cheap.

"The goal is to actually build prototype brain tissue in a dish," says Edward Boyden, a professor of neuroengineering at MIT and leader of the school's Synthetic Neurobiology Group. "We can't do experiments on live humans with anywhere near this level of resolution."

MIT and Harvard Medical School researchers have engineered a way to stack neurons to form brain cells in a dish. Image: Marcia Williams/Medical Illustrations

Initiating Cell Growth

Sculpting tissue with the new process is straightforward. First brain cells are suspended in tiny hydrogels. The gels can be as small as ten microns, about the size of a single cell, so the ultimate tissue shape can be of any conceivable or useful size. "We were happy with 100 microns," says Utkan Demirci, assistant professor of medicine and health sciences and technology at Harvard University Medical School, Brigham and Women's Hospital, and MIT. "But Ed [Boyden] said we should move this even better. We said, 'Ok, we'll spend another six months on it.'"

The gels are then sculpted photolithographically. The resultant form will determine the path of cell growth. In essence, researchers can make shapes on a photomask, which is held in place by pins over the gels. Using multiple masks, they can engineer any shape. Boyden, Demirci, and their colleagues have shown that neurons will grow from one gel to another. The brain, of course, is a many-celled thing, and the gels can host multiple cell types.

The whole system can be made for about fifty dollars.

"If you want a technology to be in widespread use, it has to be accessible. If you want tools people will use, it has to be inexpensive, scalable, and easy for people to try out," says Boyden. "I'm hoping that we described it so clearly, that they can just read it and do it."

A micro-fabricated array of multilayer, digitally specified, 3D tissue prototypes. Image: Harvard/MIT

Treating Disorders

Boyden also hopes the system will soon give us insight into a fundamental problem that has boggled brain types: How and why neurons make connections with each other. "One of the questions is: 'Can we look at how cell types emerge? Can you take stem cells and watch how they wire up?'" he asks. Stem cells from human or animal could be deposited in various points in a 3D system of hyrdogels. "Then we could watch them form connections, touch and then withdraw if it's the wrong match. It would be tantalizing to explore."

The system may solve other problems as well. People suffering from neurological disorders could have a few samples of their brains grown in digitally sculpted gel. The resulting bits of gray matter could then be treated with any number of drugs to see which they respond to best. This would save patients years of trial and error with various medications.

Already their digital tissue sculpting has answered an essential question about how the brain grows. Neurons, it turns out, are like goldfish. They grow, to some extent, to fit the size of the pond they're in. Double the size of the gel a neuron is embedded in and it will grow longer axons and dendrites, and a more complex structure.

Though creating entire brains may be something for a very far off future, engineering brain implants—for stroke victims, head trauma victims, and neurodegeneration victims, to name a few victims—may be just a synapse or two away.

Michael Abrams is an independent writer

The goal is to actually build prototype brain tissue in a dish.

Prof. Edward Boyden, MIT


March 2013

by Michael Abrams,