Growing living tissue and organs in the lab would be a life-saving trick. But replicating the complexity of an organ, by growing different types of cells in precisely the right arrangement—muscle held together with connective tissue and threaded with blood vessels, for example—is currently impossible. Researchers at MIT have taken a step toward this goal by coming up with a way to make “building blocks” containing different kinds of tissue that can be put together.

Embryonic stem cells can turn into virtually any type of cell in the body. But controlling this process, known as differentiation, is tricky. If embryonic stem cells are left to grow in a tissue-culture dish, they will differentiate more or less at random, into a mixture of different types of cells.

The MIT group, led by Ali Khademhosseini, an assistant professor in the Harvard-MIT division of Health Sciences and Technology and a recipient of a TR35 award in 2007, put embryonic stem cells into “building blocks” containing gel that encouraged the cells to turn into certain types of cell. These building blocks can then be put together, using techniques developed previously by Khademhosseini, to make more complex structures. The gel degrades and disappears as the tissue grows. Eventually, the group hopes to make cardiac tissue by stacking blocks containing cells that have turned into muscle next to blocks containing blood vessels, and so forth.

The researchers expose clusters of stem cells called embryoid bodies to a physical environment that mimics some of the cues the cells experience during embryonic development. “In an attempt to recreate that polarity, we applied microfabrication technologies to stem-cell engineering,” says Khademhosseini.

The team first puts embryoid bodies into microscale wells, which causes the cells to clump together to form spheres. Next they pour a light-sensitive hydrogel solution over the top of the cells. When this solution is exposed to light, it hardens, leaving behind a sphere of cells, half naked, half encased in a cube of gel. The process is repeated to encase the other half in a second type of gel. The result is a hydrogel block, half gelatin, half polyethylene glycol, with a sphere of embryonic stem cells inside.

Khademhosseini’s group found that within an individual embryoid body, cells on the squishier, gelatin side took a different path from cells on the polyethylene glycol side. The gelatin is easier for the cells to push into, and this affects how they grow, directing them to become blood vessels. “They completely remodel the side that’s gelatin, digging through the gel, elongating, and forming blood-vessel-like sprouts,” says Khademhosseini. These cells also express chemical markers typical of blood-vessel precursor cells, called endothelial cells. The cells on the other side differentiated in a more chaotic manner. The researchers also watched what happened when they varied the molds to create gel blocks that contained more or less gelatin.

Khademhosseini hopes to further test the effects of different hydrogels. He also plans to embed different development-stimulating chemicals within the gels. Using chemical signals to influence stem-cell differentiation is a common approach, but controlling which parts of a group of cells are exposed to which chemical signals has been difficult. Other groups have used microfluidics devices to feed different chemicals to cells. Khademhosseini believes using the hydrogel will be easier.

“This is a creative new way to guide stem cell behavior using patterned hydrogels,” says Sarah Heilshorn, assistant professor of materials science and engineering at Stanford University. She says the most innovative aspect of the work is the ability to quickly make large numbers of the cell constructs. “This approach could be applied to a broad range of other biomaterials and cell types.”

Khademhosseini’s ultimate goal is to build cardiac tissue from the bottom up. “We’d like to seed cells to pattern branching vasculature through cardiac tissue,” he says. The multimaterial gel structures, he says, “can be the modules of our self-assembling cellular structures,” he says.