3d modeling

New treatments for deadly lung diseases could be revealed through 3D modeling

Consistent with previous work, myofibroblasts do not accumulate in a soft 2d frame. Image is stained for cytoskeleton (blue), cell nuclei (yellow) and a marker for myofibroblast activation (alpha-smooth muscle actin, magenta). Image credit: Baker Lab

A bioengineered 3D model of lung tissue built by University of Michigan researchers pokes holes in decades of flat Petri dish observations of the progression of the deadly disease pulmonary fibrosis.

Brendan Baker

Brendan Baker

The causes of pulmonary fibrosis are not fully understood, but the condition is marked by scar tissue that forms inside the lungs. This scar tissue stiffens the walls of the lungs’ air sacs, called alveoli, or, in advanced stages, can completely fill the alveolar spaces. Both scenarios make it difficult to breathe and decrease the amount of oxygen entering the bloodstream. Often the disease is irreversible, eventually leading to lung failure and death.

The 3D lung connective tissue model showing lung fibroblasts (blue) within matrix fibers mimicking lung collagen (magenta).  Image credit: Baker Lab

The 3D lung connective tissue model showing lung fibroblasts (blue) within matrix fibers mimicking lung collagen (magenta). Image credit: Baker Lab

Some clinicians are concerned that critically ill patients with COVID-19 could develop a form of pulmonary fibrosis after a long stay in intensive care.

Researchers are looking for better treatments. Although they were able to find drugs that relieve symptoms or slow progression in practice, they were unable to reliably replicate these results in current 2D lab models. So they don’t understand how or why these drugs work, and they can’t always predict which compounds will make a difference. The new research from UM takes a step in this direction and clearly shows how ineffective previous approaches have been.

The team showed that in some 2D models, drugs already known to be effective in treatment do not produce test results that show their effectiveness. Their 3D tissue engineering model of fibrotic lung tissue, however, shows that these drugs work.

Before their drug trials, they first performed studies to understand how tissue stiffness leads to the appearance of myofibroblasts, cells that correlate with the progression of wound healing.

When we introduced stiffness into the 2D test environment, it activated myofibroblasts, essentially creating scar tissue.
Daniel Matera

“Even in cells from the same patient, we saw different results,” said Daniel Matera, PhD candidate and member of the research team. “When we introduced stiffness into the 2D test environment, it activated the myofibroblasts, essentially creating scar tissue. When we introduced this same type of stiffness into our 3D test environment, it prevented or slowed myofibroblast activation, stopping or slowing the creation of scar tissue.

Consistent with previous work, myofibroblasts do not accumulate in soft 2D settings (left).  In the new 3D model (right), myofibroblasts (magenta) accumulate, even under mild conditions that mimic a healthy lung.  Images are stained for the cytoskeleton (blue), cell nuclei (yellow), and a marker for myofibroblast activation (alpha-smooth muscle actin, magenta).  Image credit: Baker Lab

Consistent with previous work, myofibroblasts do not accumulate in soft 2D settings (left). In the new 3D model (right), myofibroblasts (magenta) accumulate, even under mild conditions that mimic a healthy lung. Images are stained for the cytoskeleton (blue), cell nuclei (yellow), and a marker for myofibroblast activation (alpha-smooth muscle actin, magenta). Image credit: Baker Lab

With the majority of pulmonary fibrosis research relying on 2D tests, he said, many believe that elevated lung stiffness in patients is what should be targeted with treatments. UM research indicates that targeting stiffness alone may not impede disease progression in patients, even if it works in a petri dish.

To find effective treatments, researchers first screen libraries of pharmaceutical compounds. Today, they typically do this on cells grown on flat plastic or hydrogel surfaces, but these settings often misrepresent what happens in the human body.

Timelapse of lung fibroblasts (blue) migrating and remodeling matrix fibers that mimic lung collagen (magenta) over 24 hours.  When fibroblasts are able to remodel their environment, they can transform into 3D disease-promoting myofibroblasts.  Image credit: Baker Lab

Timelapse of lung fibroblasts (blue) migrating and remodeling matrix fibers that mimic lung collagen (magenta) over 24 hours. When fibroblasts are able to remodel their environment, they can transform into 3D disease-promoting myofibroblasts. Image credit: Baker Lab

Brendon Baker, an assistant professor in UM’s Department of Biomedical Engineering, and his team took a tissue engineering approach. They reconstructed the 3D lung interstitium, or connective tissue, the home of fibroblasts and where fibrosis begins. Their goal was to understand how mechanical signals from lung tissue affect fibroblast behavior and disease progression.

“Recreating the 3D fibrous structure of the pulmonary interstitium has allowed us to confirm effective drugs that would not be identified as successes in traditional screening settings,” Baker said.

At the center of the pulmonary fibrosis mystery is the fibroblast, a cell found in the pulmonary interstitium that is crucial for healing but, paradoxically, can also promote disease progression. When activated, after injury or disease, they become myofibroblasts. Properly regulated, they play an important role in wound healing, but when improperly regulated, they can lead to chronic disease. In the case of pulmonary fibrosis, they cause stiffening of the lung tissue which interferes with breathing.

“Our lung tissue model looks and behaves similar to what we observed when imaging real lung tissue,” Baker said. “Patient cells within our model can actively stiffen, degrade, or reshape their own environment, just as they do in disease.”

The study, published in the current issue of Science Advances, is funded, in part, by the National Institutes of Health.

More information: