In a fourth-floor lab at Harvard University, Adam Feinberg is peering through a low-magnification microscope and using a scalpel to cut out triangles and rectangles from a thin polymer. What’s impossible to see with the naked eye is a one-cell-thick layer of heart tissue coating each shape. When Feinberg connects the petri dish holding the triangles and rectangles to a pacemaker, the tissue begins to rhythmically contract, and the shapes come alive–twisting, pinching, and even swimming through a solution.

The pieces of “muscular thin films” are just a few millimeters long and only 30 micro­meters thick; at first glance they resemble small worms you might find wiggling in a mud puddle. Kevin Kit Parker, the professor of biomedical engineering who heads the Harvard lab, jokes that he’s planning to retire to the South, where he was raised, and sell them as customizable lures in a bait shop.

But the experiment has entirely serious implications. Eventually, the patches of twitching tissue could be used as actuators for tiny robotic devices implanted in the body. The muscle cells would be fueled by sugar in the bloodstream and maintained by the same repair mechanisms that keep the heart pumping. Parker says the muscle­-coated film could also be used to regenerate tissue damaged in heart attacks. But such applications are quite a way off, he says. In the nearer term, the devices could be used to help researchers monitor how experimental medicines change the behavior of heart muscle.

Printing Tissue
This is not the first time researchers have grown beating heart muscle in a dish. But Parker and Feinberg, a postdoctoral researcher in Parker’s lab, have found ways to make the tissues far more powerful, contracting with the same strength as natural heart tissue.

The manufacture of the devices begins with a biological printing technique, developed by Harvard chemists, that can deposit proteins in microscopic patterns on various surfaces. Parker and Feinberg use the method to precisely organize the heart cells into working tissue.

The process looks unremarkable. Working in a sterile laboratory hood, Feinberg arranges a few chunks of clear silicone rubber in a petri dish. The chunks are stamps patterned with an array of microscopic lines. The pattern was created by molding the stamps to a wafer of silicon etched using the same techniques that produce microchips. Onto each stamp Feinberg squirts a clear “ink” that contains a common protein called fibronectin. As the stamp dries, a thin layer of the protein forms. Holding a stamp with a pair of forceps, Feinberg presses it onto a round, silicone­-coated glass coverslip, transferring proteins from the raised portion of the microscopic pattern to the silicone film.

With the protein patterns stamped out and ready, Feinberg immerses the coverslip in a solution of young, still developing heart cells harvested from rats. The cells begin to adhere to the fibronectin, forming orderly lines. Feinberg then puts the cells and the protein-patterned coverslip, still immersed in the solution, into a body-­temperature incubator. Over the next few days, the lines of fibronectin guide the cells’ organization and further development. Long, fiberlike contractile units begin to form, guided by the cells so that they line up parallel to the lines of protein. If they weren’t aligned this way, the cells would fight against each other as they contract rather than pulling in the same direction. The aligned cells, however, all contract along the same axis, much the way they do in natural heart tissue.

When Feinberg removes the newly grown tissue from the incubator, it and the film of silicone it’s printed on are immobilized by the rigid glass coverslip. But as they cool, a temperature-sensitive glue that holds the silicone to the glass begins to dissolve. Feinberg has just a few minutes to cut out shapes before the silicone and tissue float free. Once they do, the heart tissue can contract, making the film to which it’s anchored start bending and twisting.

So far Feinberg has made rudimentary pumps, twisting actuators, pincers, a device that slowly swims, and another that walks along the bottom of a petri dish. A long rectangular strip, cut from the film so that the lines of cells run along its length, curls up with each contraction. Another rectangle, cut at a slight angle to the cells, coils up into a corkscrew. The narrow “tail” of a triangular piece propels the shape through the solution. The behavior of these devices can be controlled like that of a natural heart: with a pacemaker. Feinberg hooks electrical leads to the small dish holding the devices. Low-voltage bursts of electricity travel through the solution, signaling the muscle to contract.

Muscles on Drugs
A practical way to measure the effect of drugs on heart tissue is to determine how strongly treated tissue can contract. Thus, the device likely to be most useful in the short term is also one of the simplest: a long rectangular strip of tissue that bends slightly with each pulse of electricity. These devices could be used both to screen drugs meant to act on the heart and to identify drugs that may adversely affect the heart.

Because the mechanical properties of silicone are well known, it’s possible to determine exactly how much force the heart tissue is exerting by measuring how much the strip bends. If a change is observed in the amount of force the cells can exert, it’s a sign that a drug is having an effect. Parker envisions a testing system of small wells, each containing a strip of silicone and heart muscle. Such a system could be used to measure the effects of different compounds, or different concentrations of the same compound, on the heart tissue’s ability to function. The system could even be automated; Feinberg has already developed software that analyzes video of the strips and calculates changes in the amount of force the tissue exerts.

So far, the researchers have used only rat cells. Eventually, they hope to make screening tools with human cells, perhaps by first growing stem cells and then coaxing them to develop into heart cells. They also hope to make similar systems with muscle cells that line blood vessels–to test hypertension drugs, for example. For other applications, the devices will have to be made either smaller (for implantable robots) or larger (for patches that help heal damaged hearts).

Ultimately, the key to the technology may be its simplicity, which could make it easy to adapt to a range of applications. As Parker says, “We have dummy-proofed this tech­nology so that it is easy to learn, easy to do, and eventually, easy to deploy in the clinic.”