Biomedical engineers have built many types of human organs in the lab, but they've lagged on lung tissue--until now. Two new studies have used very different approaches to do the job.

One team has grown lung cells that performed their gas-exchange functions when transplanted into living rats. The study provides proof of principle that such regenerated tissue may one day be used to treat patients with serious lung disorders.

The other has built a microfluidic chip that mimics lung function, resulting in a biologically relevant model for testing medicines for lung disorders or conducting toxicity screens for nanoparticles.

"Lungs don't heal themselves, and when we transplant lungs, patients tend to do very badly," says Laura Niklason, a biomedical engineer at Yale University in New Haven, Connecticut. Her group wanted to make lung tissue from the body's own cells, to prevent rejection when transplanted.

They decided to grow the cells on the lung's extracellular architecture rather than on synthetic cell matrices--the approach most researchers have focused on until now.

Training for cells

Niklason's team first gently removed the lung cells from their support architecture--made mostly of collagen--leaving it intact. They then added eight or nine cell types back onto the matrix and cultured them for eight days.

To the team's surprise, the natural extracellular matrix directed the cells to their correct locations, essentially "training" them to function appropriately. Finally, the researchers transplanted the tissue into rats.

Much like Niklason's study, published online June 24 in Science, two other recent studies used such decellularized matrices rather than synthetic ones. What has never been done before, however, says Peter Lelkes, a tissue engineer at Drexel University in Philadelphia, Pennsylvania, is transplanting the bioengineered lung tissue into a living organism.

"We were actually kind of surprised at how well they worked," Niklason says. The transplants functioned for about two hours, fulfilling the lungs' key function of exchanging oxygen for carbon dioxide.

In practice, however, lung cells--especially from older, ill patients--won't grow well enough in culture, but will have to be produced from stem cells or induced pluripotent stem (iPS) cells, she says.

"What we've really done is developed the technological and scientific underpinnings so that we can bring stem-cell biology to bear on this problem."

Angela Panoskaltsis-Mortari, who studies lung injury and repair at the University of Minnesota, Minneapolis, agrees. Panoskaltsis-Mortari, an author on one of the other recent lung-engineering studies using a similar matrix, and her group presented data at a conference last month showing that iPS cells can differentiate into a key type of lung cell when grown on decellularized matrices.

"It's a really ideal model" for studying lung differentiation, lung infection and related topics, she says.

Mechanical mimicry

In the second study in Science, researchers from Harvard University in Boston, Massachusetts, created a chip 1 to 2 centimeters long in which a 1 millimeter-wide channel, coated with human lung cells on the inside and overlaid with human blood capillaries on the outside, mimicked the air sacs, or alveoli, of the lungs.

The idea, says biomedical engineer and study leader Donald Ingber, was to create a way to do high-throughput screening--used by the pharmaceutical industry to screen large numbers of molecules--but in a system that's biologically closer to human tissue than to a petri dish.

What is most impressive about the chip, says Michael Shuler, a biomedical engineer at Cornell University in Ithaca, New York, is that it mimicked not just the physiology of the lung, but also the mechanical forces that act on lung cells as the chest expands and contracts with each breath.

The group also tested the chip as a system for determining how different types of nanoparticles penetrate the lungs, and how bacteria enter the lungs to cause infection. Ingber says he and his colleagues are now contacting pharmaceutical companies to find out how the device can best be used.

"As a system by itself, it could be used to test drugs very directly," says Shuler. Only one out of ten drugs in development makes it successfully through clinical trials, he adds.

"Maybe by having a more authentic system, you change the odds" to more like three out of ten, he says. "It doesn't have to be perfect to be valuable."