The lub-dub of a healthy heart
Developmental biology guides efforts to “grow” replacement heart valves
The containers on the conference room table are the type you might store leftovers in. But these don’t hold last night’s spaghetti or week-old casserole from the church potluck. They hold human hearts.
He proceeds to describe how the heart works as a pump, opening it to reveal the heart’s right and left chambers, its muscular walls, and the valves that control blood flow. Then Barnett places the first heart into the tentative gloved hands of the man closest to him, to pass around for a closer look.
Barnett reaches for another heart, pointing out the blood vessels and the tough remains of an atherosclerotic plaque – “the most dangerous thing in the western world,” he says.
He saves for last the heart that intrigues him most. This heart has been through a lot. The valves at the openings of both the pulmonary artery and the aorta – the large blood vessels that send blood to the lungs and to the rest of the body – have been replaced. The sewn-in aortic valve is a natural tissue valve; the pulmonary valve is artificial.
“This is what we want to do away with,” Barnett says, pointing at the artificial valve.
By now, the group is eagerly examining the hearts and asking questions – even the man who originally sat down is back on his feet. These on-air radio personalities and producers will be conducting the Children’s Miracle Network radiothon in the coming weeks; they are looking at hearts in a Vanderbilt conference room to learn a bit about how biomedical research leads to new therapies.
Barnett, who also is vice chair of Pharmacology, is a good teacher. His hands-on heart tutorial evolves naturally into a discussion of his quest to discover the genes and signaling pathways that build the developing heart and its blood vessels – research that could make it possible to grow replacement heart valves in the laboratory, from a patient’s own stem cells.
“Here’s the fantasy,” Barnett says. “If you need a valve, we take your circulating or bone marrow-derived stem cells, have tissue engineers sculpt the right shape and size matrix, grow the cells on the matrix, and give the complete heart valve to the surgeon.
“Wouldn’t that be amazing?”
An outrageous hypothesis
A preserved human heart (also pictured on page 23) displays its history: a variation of the extensive Ross procedure, in which surgeons replaced a diseased aortic valve with the heart’s own pulmonary valve. An artificial valve (shown here) was then stitched in its place
The “lub” happens when the mitral and tricuspid valves – the valves separating the atria and ventricles – close. The “dub” corresponds to the closing of the aortic and pulmonary valves.
Valve failure – because of a developmental defect or disease – forces the heart to work harder to compensate for the defective blood flow, leading in many cases to congestive heart failure.
About one in 100 children is born with a congenital heart defect, the leading cause of death in the first year of life.
“Depending on the numbers you use, somewhere between 60 and 80 percent of these kids have an abnormal valve,” says Scott Baldwin, M.D., chief of Pediatric Cardiology at Vanderbilt who, with Barnett, is leading the charge to grow replacement heart valves in the lab. Children with valve defects often require multiple valve replacement surgeries as they grow.
“On the flip side, up to 4 percent of the population over the age of 60 will need to have a valve replaced because it’s calcified and thickened,” Baldwin says. More than 100,000 valve replacement surgeries are performed each year in the United States, according to the American Heart Association.
“Heart valve problems have become epidemic; it’s an important issue,” Baldwin says.
So why grow valves in the lab?
Existing options for valve replacements are not ideal, Baldwin and Barnett argue.
Artificial (mechanical) valves are durable, but patients require lifelong blood-thinning therapy. Tissue valves – from pig, cow or human hearts, sometimes with artificial parts – don’t usually necessitate blood-thinning treatment, but they will only last a decade or so.
Enter SysCODE (Systems-based Consortium for Organ Design and Engineering), an interdisciplinary group that will work toward growing heart valves – and also teeth and pancreatic islets – in the lab.
Led by Richard Maas, M.D., Ph.D., chief of the Division of Genetics at Brigham and Women’s Hospital in Boston, SysCODE was awarded a five-year, $24 million grant last year as part of a National Institutes of Health “Roadmap” initiative that is designed to speed the movement of scientific discoveries from the bench to the bedside.
The premise of the SysCODE program, Baldwin explains, is that development follows a “blueprint” for forming complicated organs from a single cell type. It’s up to the investigators to decipher this blueprint – determine which genes are the essential ones – and use that information to push the appropriate stem cell populations to form heart valves, teeth or pancreas.
Starting with embryonic stem cells from mice, “we’re going to figure out every gene involved in each of those developmental programs,” Baldwin says. “Ultimately, I would like to be able to take patients’ own stem cells and give them back a valve.
“It’s an outrageous hypothesis,” he adds, laughing.
The investigators already know a lot about the cells that will multiply, transform and become heart valves. They are a special subset of the endocardial cells that will line the heart.
Let’s back up.
When the heart initially develops during the third week of human embryogenesis, it is a simple tube with two epithelial layers of cells: an inner endocardium, which will form the inner lining of the heart, and an outer epimyocardium, which will become the heart muscles that will pump for the lifetime of the individual.
In between the two layers is extracellular matrix, gelatinous material void of cells called “cardiac jelly.” At the sites where the heart valves will take shape, Barnett explains, the tube constricts and the cardiac jelly expands, becoming a bulge called the cushion.
“If you look at a heart at this stage of development the bulge is already functioning as a valve,” Barnett says. “It’s very resilient, so when the heart pumps, the blood moves through and then this bulge snaps back into place to prevent blood from flowing backwards.”
Next, a signal (or signals), most likely made by muscle cells, causes some of the endocardial cells to change from epithelial-type cells to connective tissue mesenchymal-type cells, migrate into the cardiac jelly and populate it. This change in cell type, known as “epithelial-mesenchymal transformation” (EMT), is a crucial step in the formation of each of the tissues – heart valves, teeth and pancreas – that the SysCODE consortium will attempt to build in the lab.
Barnett’s group and others have extensively studied the EMT process in valve-forming cells in embryonic chick hearts.
His team removes the cushion region from a 2-day-old developing chick embryo, grows it in culture, and adds or subtracts growth factors to evaluate which factors affect cell invasion and proliferation. The group also injects viruses into chick embryo hearts, which are the size of a comma on this page, in order to change gene expression and assess the impact on transformation.
“We know a lot about the transformation process,” Barnett says, “but then next we talk about magic happening. You’ve got the cushion full of cells, and by some combination of genetic and hemodynamic forces, that cushion gets remodeled into what we call a heart valve.
“That’s where we’re heading with SysCODE, is to understand – and eventually replicate in vitro – that remodeling process.”
Heart in a Petri dish
Barnett, who earned his Ph.D. in Pharmacology from Vanderbilt in 1986, remembers exactly when he fell in love with the idea of studying the heart. He was visiting Harvard cardiovascular researcher Jonas Galper, M.D., Ph.D., to talk about a postdoctoral fellowship when Galper suggested they take a look at the embryonic chick heart cells the group was just beginning to grow in culture.
“When I saw those beating heart cells, I knew immediately that I would be working with Jonas,” Barnett recalls. “Here was a cell that was beating in culture, that had a biological response that you could measure – a cell that you could use to really ask questions about how signaling molecules affect and regulate the biology.”
Barnett and Galper followed gene expression and protein changes at a particular stage in chick heart development to discover how the heart muscle begins to respond to hormones that regulate heart rate. Before day 3, drugs that normally change the heart rate have no effect; in the next day or so, “those systems come online, and you’ve got a heart that behaves like an adult heart,” Barnett says.
As Barnett was finishing up his postdoctoral studies, a chance elevator conversation with Maas, a friend from graduate school at Vanderbilt, led to a collaboration between the two that kindled Barnett’s interest in how the heart’s structure forms.
He began to pursue the role of TGF-beta, a widely expressed growth factor involved in cell proliferation and maturation. It was first described in the early 1980s by Mayo Clinic researchers led by Harold Moses, M.D., who later directed the Vanderbilt-Ingram Cancer Center.
TGF-beta also is one of the factors that Barnett and Galper identified as key to turning on the hormone response in chick heart muscle. At Vanderbilt, Barnett and colleagues found that a particular TGF-beta receptor – the type III receptor – is essential for the transformation of endocardial cells in early heart valve formation.
Interestingly, the chromosomal region that is home to the type III TGF-beta receptor gene has been linked to congenital defects in the valves and dividing walls of the heart. Other scientists are looking for the variants that cause the defects.
Barnett and colleagues continue to tease apart the complexities of TGF-beta receptor signaling. They recently discovered that the type III TGF-beta receptor also has a role in the development of coronary blood vessels.
“It’s a recurring theme in developmental biology that nature uses the same, or similar, molecules over and over again in slightly different contexts,” Barnett says.
Bit of serendipity
Baldwin traces his research path back to a lecture at a Society for Pediatric Research meeting that he attended during his residency. At the meeting, Merton Bernfield, M.D., a Harvard neonatologist and pioneer on studies of the extracellular matrix who died in 2002, talked about how organs take shape.
“I was mesmerized,” Baldwin recalls. “I had already committed to a cardiology fellowship, and I remember sitting there thinking ‘I want to know how the extracellular matrix influences heart development.’”
During his fellowship research at the University of Iowa, Baldwin published the first paper demonstrating that a component of the extracellular matrix, hyaluronic acid, is important for heart development. And he became intrigued with the question of what patterns the heart – how the single tube loops and twists into an organ with chambers and valves. He suspected that the endocardial lining was involved in laying down the template for the heart.
At the time though – the late ’80s – it wasn’t possibly to identify endocardial cells, or even their endothelial cell precursors, in the embryo, Baldwin notes. Undeterred, he joined a lab at the Wistar Institute in Philadelphia, where over the next several years he cloned a mouse gene that identified endothelial cells.
Then came a bit of serendipity. In the late 1990s, Harvard immunologist Laurie Glimcher, M.D., knocked out the gene for a transcription factor in mice to study its effect on immunity and found that the mouse embryos died in utero. She suspected a heart defect and asked Baldwin to take a look. It turned out that the mice didn’t form aortic or pulmonary valves.
Since that discovery, Baldwin and colleagues have found that the gene for this “nuclear factor of activated T cells” (NFATc1) is expressed not only in the subset of endothelium that will become endocardium, but also in the particular endocardial cells that will become the heart valve.
By fusing the gene for a fluorescent protein to the portion of the genome that regulates expression of the NFATc1 gene and then inserting this “marker” into mouse embryos, the investigators can now track – by their glow – cells that are destined to make up the heart valve, isolate them for in vitro studies, and even use the system to understand what factors are essential for heart valve formation.
“So we think we’ve got the building blocks; now we’ve got to figure out how to put them together and get them to do what they’re supposed to do,” Baldwin says.
Part of the “putting them together” means getting the matrix right.
If you look at a heart valve, it’s mostly not cells; it’s mostly matrix,” Barnett says.
“We know a lot about the cells – we know one when we see one, and we know a lot about the factors that make those cells transform,” he continues. “But as far as what’s in that matrix, I could fall in a bucket of it tomorrow and not really know what I was in. It’s the missing component right now.”
Richard Caprioli, Ph.D., who directs the Vanderbilt Mass Spectrometry Research Center, will lead efforts by the SysCODE consortium to de-mystify the matrix by identifying the proteins in the cardiac jelly that are important in valve development.
Already, tissue engineers in Boston are making “gel substrates” of hyaluronic acid – the matrix component Baldwin identified as important for heart development. “As we learn what other components are in the matrix, and how they’re organized, our Boston colleagues can make and incorporate those things into the gel,” Barnett says, to create the best “scaffolding” for a heart valve.
Ultimately, to grow a valve in the lab may take more than the right scaffolding and the right cells, Barnett and Baldwin acknowledge. The remodeling that takes place after the cells are in the cushion matrix in vivo may require pulsatile blood flow, and the investigators are in conversation with the tissue engineers about how to potentially replicate that flow in vitro.
The parts will come together, these investigators say.
“It sounds like science fiction, but it will happen in my lifetime, there’s no question,” Baldwin says. “There’s no inherent design limitation; the only thing we don’t have is all the information, which is what we need to get.”
The investigators are also optimistic that SysCODE’s unraveling of the developmental biology “programs” for heart valves, teeth and pancreas will turn up common signaling pathways and explain why it looks like valves turn into bone – become stiff and calcified – when they’re diseased.
“We think that maybe when the valve is injured, it reactivates developmental programs to try to repair itself; only now these are not appropriate,” Baldwin says. “If we understand the developmental program, I think we’re also going to figure out what the pathological program is … and if you can find something to prevent aortic valve calcification as people age, that’s going to have a huge world health impact.”
The radio folks leaving the conference room seem convinced. “That was cool,” says one. “It’s incredible what these researchers can do.”