One alumnus’ work is bringing us closer to a cure for Type 1 Diabetes – and that’s thanks in part to a helpful UWM graduate student.

Dr. Martin Hessner graduated from UWM in 1990 with his PhD in Microbiology. He is now a professor with the Department of Pediatrics and the Director of the Max McGee Research Center for Juvenile Diabetes at the Medical College of Wisconsin. He is currently working on a genomics approach that indicates the likelihood that someone will develop Type 1 Diabetes.

Type 1 Diabetes, also known as juvenile diabetes, is an auto-immune disease in which the body attacks its own pancreatic beta cells. These cells produce insulin, a hormone needed by the body to process sugars. Type 1 Diabetes, if not properly managed, can lead to heart or blood vessel disease, kidney and nerve damage, as well as other serious complications. Untreated, it can be fatal.

“The most relevant tissues needed for study, such as pancreatic biopsies, are not accessible. Therefore, it is difficult to directly study Type 1 diabetes in humans,” Hessner said. “We have developed an alternative approach where we draw plasma or serum – the liquid part of blood – from the person to be studied. We use the plasma to elicit a response in healthy leukocytes taken from a healthy unrelated donor. We ask the question, what immune factors are in the plasma of the person we want to test that will turn on genes in those leukocytes?”

Hessner and his team have worked with hundreds of families affected by Type 1 Diabetes, in some cases for up to a decade, to build up a sample bank of blood collected from both affected and unaffected children. Over time, they can look and see how the immune state measured in the blood samples changes as the disease does, or doesn’t, progress inside of the subjects Hessner and his team follow. During disease progression, inflammatory genes tend to be expressed, or become more active, in subjects progressing to Type 1 Diabetes, while other genes, called anti-inflammatory genes, are expressed more in non-progressors.

“What we find is that kids who are progressing to Type 1 Diabetes, if you look at them over time, they drive (gene) expression with their plasma, and the reporter cells exhibit an inflammatory response. You can see a crescendo in that inflammatory response as they get nearer and nearer to clinical onset (of Type 1 Diabetes),” Hessner said.

To examine the blood, Hessner and his team use a microarray to measure the response in up to 54,000 genes at once. That’s an overwhelming amount of data to sort. In order to investigate additional ways to prioritize the genes induced by plasma of patients with Type 1 Diabetes versus subjects without it, Hessner and his worked with UWM and Mathematics PhD student Sami Cheong.

Cheong works with statistical modeling and algorithms, and she already had a background in Biology as a mentor in the Undergraduate Research Fellowships in Biology and Mathematics program. In 2013, she was asked to help Hessner and his team explore mathematical models that could be applied their data.

“One of my tasks was to reduce the variables (the massive numbers of genes induced by the plasma samples). We were able to reduce it to about 700 genes. From there, I used some statistical algorithms to make the list even smaller. We used various statistical algorithms, such as cluster analysis and machine learning, as well as different exploratory methods, to investigate the underlying variability of the data. We were able to use the list of genes to compare the individual trend for each person,” Cheong said. “We calculated a score from the reduced variable size. We basically plotted for every time-series data that we had and looked at what the trend is for everybody. They have some idea from that approach that if someone has an upward trend, they are likely to develop Type 1 Diabetes. If they have downward trend, they are not likely to develop Diabetes.”

Much of Cheong’s work was trial-and-error as she put together various algorithms to try and see which model would work best to distinguish between the healthy blood samples and the samples with Type 1 Diabetes. It took a lot of time, but eventually, Cheong said, “I was able to find the best algorithm to separate the two groups. I was able to identify genes that contribute to the separation.”

Now, two years later, Hessner is using Cheong’s algorithm and his team’s research in clinical trials.

“We can use Sami’s algorithm and others to study trends in the immune state of an individual and make predictions whether they’re progressing or non-progressing. We can also use these same algorithms see the effect of therapeutic intervention,” Hessner said. “We’re looking at different clinical trials to see how the inflammatory state associated with Type 1 is altered when different drugs are tested.”

Hessner hopes that this research can lead to better treatment options for Type 1 Diabetes, and eventually, even a cure.