Excelling Together - Longstanding collaboration drives duo's diabetes research

The islets of Langerhans sounds like an exotic destination — a South Sea archipelago with swaying palm trees, white sands and pristine waters.

Though they have nothing to do with geography, the islets of Langerhans are indeed exotic and mysterious, scientifically speaking. The term refers to clusters consisting of several types of cells in the pancreas — alpha, beta, delta and gamma. Beta, the most common cell type, produces and releases insulin. Diabetes mellitus is caused when beta cells fail to secrete enough insulin into the blood or when a sizable fraction of beta cells die.

Florida State University researchers Michael Roper and Richard Bertram have combined the power of their individual disciplines in a collaboration that explores the complexities of the pancreas, the islets and how these interact with the liver. The pair’s research could offer insights into how to reactivate malfunctioning or shutdown beta cells and have far-reaching implications for the more than 34 million Americans living with diabetes.

Complementary colleagues

Roper, who earned his Ph.D. from the University of Florida in 2003 and came to FSU in 2006 after completing postdoc work at the University of Virginia, heads the Roper Lab, which has a particular focus on developing methods to measure the functions of the islets of Langerhans. Bertram, director of biomathematics in FSU’s Department of Mathematics and a member of the Institute of Molecular Biophysics, has been developing predictive mathematical models to further islets research since 1993, when he earned his doctorate at FSU and went to work for the National Institutes of Health. He returned to FSU in 1999.

“Bertram and his colleagues are world leaders in developing mathematical models of how the pancreas releases insulin,” Roper said. “I contacted him within a few weeks of starting at FSU. We had coffee and hit it off.”

That first meeting launched both an ongoing scientific partnership and a friendship.

“I had an idea for a project based on work I’d done previously and wanted to test it but needed the perfect setting,” Bertram said.

He found such a setting in Roper’s lab, where Roper and his colleagues test signaling processes like the one that triggers the pancreas to release insulin.

“Insulin comes out in pulses every 5 minutes or so. How that happens is still unclear,” Roper said. “We’ve spent the last 10 to 12 years trying to figure out how these impulses may be brought about. In diabetes, you lose these pulses.”

The human pancreas.

Precision models

To test Bertram’s theoretical models in the lab, Roper uses specially modified microfluidic devices — credit card-sized pieces of glass etched to produce microscopic channels that can house biological material such as islet cells — to stimulate the cells with precise doses of glucose.

“The channels in the device act almost like blood vessels, like the environment in the pancreas,” Roper said. “We can measure insulin more precisely and test Bertram’s models.”

When developing a biomathematical model such as those used in islets research, Bertram likes to start small and simple.

“My philosophy is to only put in the things you have to put in,” he said. “The first model of the electrical activity of pancreatic beta cells was developed in the mid-1980s (not by me) and had four differential equations. That was sufficient to explain a lot.

“Over the years, the number of equations has fluctuated as models develop. Some of my simplest models have four equations, others have 12 to 14. The more complicated models have 70 to 80 parameters.”

Collaborative catalyst

Computational models play an increasingly important role in physical and biological sciences, and developing a great computational model can aid scientists in hypothesis testing, finding or predicting patterns in data or behavior, and spark new ideas or approaches for experiments, said Joel Adablah, a fifth-year graduate student in Roper’s lab.

“Although pancreatic islet biology was the common research interest, the two research labs involved had very different expertise,” he explained. “This is where the strength of the collaboration lies.”

Existing computational studies may inspire the innovation of new analytical tools, which then allow for investigation of previously unexplored aspects of islet biology. Once experimental findings are developed, computer modeling can confirm results and help scientists extrapolate further understanding of how the pancreas functions. 

“You can observe things with the model and come up with predictions and then test them experimentally. Science is about testing hypotheses,” Bertram said. “When you do computer simulations, you often learn things intuition doesn’t tell you. If you consider things long enough, you can generate good theories, but computer simulations sometimes yield totally unexpected results and those counterintuitive things usually have the biggest impact.”