Breakthroughs begin here: Revolutionizing diabetes research
UF-developed biorepository, artificial intelligence tools lead to continued innovations and therapies
July 1, 2026 — A German pathologist peered into a microscope in 1902 at a slice of tissue from a patient who had died of a disease that would later be called Type 1 diabetes. He saw white blood cells infiltrating clusters of pancreatic cells whose purpose remained an enigma.
It was the first foundational clue that Type 1 diabetes was an autoimmune disease, where the body’s immune system mistakenly attacks its own healthy tissue. Those mystery cells produced insulin and were destroyed by the disease.
But researchers didn’t fully grasp any of that; the limits of primitive technology at the infancy of immunology meant scientists were essentially stumbling in the dark.
Todd Brusko, Ph.D., leads a team of researchers as director of the UF Diabetes Institute in advancing discoveries for diabetes therapies.
Now, University of Florida College of Medicine and UF Diabetes Institute researchers are at the forefront of a revolution in diabetes research. Among public universities, UF is the No. 1 recipient of National Institutes of Health training funding for Type 1 diabetes over the past three years — and the only one in the Southeast.
Using the latest technology, advanced imaging techniques and computational power, UF investigators have unprecedented insight into the onset and progression of Type 1 diabetes. Along the way, researchers have discovered numerous new targets for potential therapies to interrupt and slow disease progression.
“This didn’t happen overnight,” said Todd Brusko, Ph.D., a leading researcher in the field and director of the UF institute. “Within the last five years, the technologies have transformed our capacity to look with single-cell resolution at the actual target tissue itself. We’ve been getting very good at treating diabetes from a technological standpoint, and now we’re starting to make headway in understanding the underlying processes driving autoimmunity.”
A repository that forever changed diabetes research
UF Diabetes Institute founding director Mark Atkinson, Ph.D., established the repository now used by investigators worldwide for diabetes research.
The crucial first step in this explosion of knowledge was the 2007 creation of the UF-based Network for Pancreatic Organ donors with Diabetes, or nPOD, the world’s largest biorepository of pancreatic tissue used in Type 1 diabetes.
Mark Atkinson, Ph.D., a UF College of Medicine eminent scholar and former longtime director of the UF Diabetes Institute, along with a team of UF researchers, established the repository now used by investigators worldwide for diabetes research.
“The nPOD team has been collecting samples for over two decades to find the most important signals of disease,” said Brusko, also a professor in the UF College of Medicine Department of Pathology, Immunology and Laboratory Medicine. “The goal has always been to inform new therapies. nPOD helped make that transformation possible. And then, within the last five years, the technologies have really transformed our capacity to examine target tissue with single-cell resolution. It’s going to be absolutely transformative for how we think about what we target.”
The ability to study high-quality samples of human pancreatic tissue from donors at varying stages of the disease enabled breakthroughs by scientists in Gainesville and worldwide. The research examples below, for example, relied on pancreatic tissue obtained from nPOD.
“One of the things that we at nPOD tried to do was to support the use of emerging technologies,” said Atkinson, who, like Brusko, is a professor in the UF College of Medicine’s Department of Pathology, Immunology and Laboratory Medicine. “In previous decades, we were throwing Hail Mary passes at Type 1 diabetes seeking prevention or cure, trying this, trying that, hundreds of clinical trials without knowing what causes the disease.”
Technological advances allow a more targeted approach.
Hampering autoimmune attack on insulin-producing cells
Edward Phelps, Ph.D.
An April study involving UF researchers from the College of Medicine and Herbert Wertheim College of Engineering demonstrated that an experimental drug developed by the United Kingdom-based biotech company Immunocore inhibits T cells that attack insulin-producing cells in the pancreas.
In experiments, insulin production dropped by more than half. Using the drug to block T cell-mediated killing of beta cells, insulin levels remained normal, slowing the progression of Type 1 diabetes.
Edward Phelps, Ph.D., a member of the Diabetes Institute and senior study author, said the drug helped prevent the loss of insulin secretion in an experimental model using materials from nPOD. Immunocore plans to conduct early-phase clinical trials initially in Europe.
Development of drugs to slow the progression of Type 1 diabetes has largely relied on broad immune suppression. A more targeted approach, like the one the study explored, could cause fewer deleterious side effects to immune function and act only at the site of the autoimmune attack. Phelps noted the drug essentially tells the T cells to stand down.
“The beauty of this drug is that it is tailored and targeted at the beta cells that produce insulin,” said Phelps, an associate professor in engineering with an affiliate appointment in medicine’s pathology department. “It basically prevents the loss of insulin secretion.”
Mapping hidden immune signals of Type 1 diabetes
UF College of Medicine researchers working with the University of Pennsylvania, published a study in March that provided the most detailed map yet of gene activity in the pancreas of individuals with Type 1 diabetes. This will help scientists pinpoint the biological signals driving the disease.
A method called spatial transcriptomics enabled investigators to capture gene expression signatures from thousands of spots on thin pancreatic tissue slices from nPOD. The map showed increased activity in a gene associated with inflammation commonly targeted in rheumatoid arthritis, which scientists believe also contributes to Type 1 diabetes.
“You’re looking at a map, almost like a topographical map of an organ or tissue and saying, in this particular ZIP code or region, these are the genes that are highly expressed,” Brusko said.
The finding makes the gene a potential target for future therapies. The gene map can also help determine if other drug therapies might prove effective.
An AI-powered potential route to early diabetes detection
Researchers at the Diabetes Institute analyzed blood samples from thousands of people in a February study to identify hidden immune-system patterns linked to Type 1 diabetes, focusing on T cells that attack insulin-producing cells. Each T cell carries a type of molecular fingerprint called a T-cell receptor. In those who develop Type 1 disease, T-cell receptors recognize components of beta cells, such as insulin.
Using advanced genetic sequencing and artificial intelligence, the scientists identified recurring receptor patterns in T cells that were more common in people with Type 1 diabetes and in those at high risk for developing it. Some of the same immune cells were also found in lymph nodes near the pancreas.
The work could eventually help researchers develop blood tests that detect the autoimmune attack sooner, possibly before symptoms appear, allowing for earlier therapeutic interventions and ways to better track how the disease progresses and whether experimental treatments are working.
“You can think of an individual’s immune system as a USB key that carries the memory of prior exposures over a lifetime,” Brusko said. “This code can be linked to infections we encounter throughout life, but it also has the potential to show the immune response in those who develop autoimmunity. AI is now helping us to decode and track these immune signals in those who develop Type 1 diabetes.”
