Decades of painstaking laboratory work has led a University of Colorado School of Medicine team to a fundamental discovery pointing to the roots of Type 1 diabetes and, possibly, other autoimmune diseases. They identified what appears to be the molecular lure that leads T cells – which normally protect the body from invaders – to mistakenly attack and destroy the pancreas’s insulin-producing cells. The result: 1.25 million people in the United States who must rely on insulin pumps or injections for the rest of their lives.

Thomas Delong, PhD, and Kathryn Haskins, PhD, led a CU Department of Immunology & Microbiology team in the effort, the results of which were recently published in the prestigious journal Science.

Delong, a biochemist and himself a Type 1 diabetic, arrived at CU from the University of Erlangen-Nuremberg in Germany a decade ago. He came to apply his expertise in chemistry to an effort Haskins had started in the late 1980s to study T cell clones from mice that spontaneously develop autoimmune diabetes. The aim all along was to find out what antigen or antigens in the pancreas’s insulin-producing beta cells were activating T cells and triggering the autoimmune attack that leads to Type 1 diabetes.

Understanding that trigger would give researchers an invaluable tool for developing new ways of testing for Type 1 diabetes long before symptoms appear and creating therapies to stop the disease’s progression before the body’s own immune system wipes out the pancreas’s ability to produce insulin.

Tough problem

When Delong arrived in 2006, it was thought that the antigens that attracted T cells were proteins neatly encoded by some straightforward section of DNA. The most obvious of these would have been insulin itself, which, given Type 1 diabetes’ attack on insulin-producing cells, might logically play the role of blood in shark-infested waters.

It’s a neat comparison that unfortunately isn’t quite so pat. “This is a very difficult problem,” Haskins said. “T cells don’t respond to a protein as a whole protein – that protein has to be broken up into pieces.”

Cells routinely break apart, reconstitute, tweak and customize proteins as they roll off their molecular assembly lines in a process called post-translational modification, Haskins explained. She and Delong hypothesized that whatever rogue antigen was attracting T cells in the pancreas was probably a function of the protein modification process going awry without leaving an obvious blueprint behind.

“It’s a black box,” Delong said. “You don’t know what the T cell is actually seeing.”

Pep rally

And so Delong focused on chunks of proteins called peptides. He synthesized them, purified them, and tested how attractive they were to a panel of lab-grown mouse T cell clones. By 2010, he had found T cells to be weakly attracted to a peptide called WE14, a short chain of amino acids lopped off a protein called chromogranin A that is abundant in beta cells produced by the pancreas.

Delong wasn’t especially excited. “The T cells responded poorly – it was a lousy antigen,” he said.

But at least T cells did respond. The discovery of something even remotely magnetic to rogue pancreatic T cells was enough to land the work in Nature Immunology. Delong and Haskins then focused on possible combinations of peptides that pancreas cells might put together, Lego-like, through post-translational modification.

“We were looking for the needle in the haystack,” Delong said. But it was really nothing so simple as that, he added. With WE14, “We knew a piece of the needle, but we suspected the needle was somewhat modified. It was painted like hay.”

Trial and error

Delong proceeded to invent different combinations of peptides that might combine with WE14. The idea was to create an antigen that pancreas cells could conceivably be synthesizing that attracted T cell friendly fire. He called them hybrid insulin peptides, or HIPs, and the library of 187 of them he created itself broke new ground in the field of immunology, Haskins says. Delong tested many combinations with cloned mouse T cells, waited 24 hours, and looked for gamma interferon, an inflammatory protein that T cells release when the right antigen whets their appetite for destruction.

Insulin, it turned out, did play a role in that – but in the form of a fragment of a peptide cleaved from insulin cells. The fragment chemically bonded with the WE14 amino acid chain, and the resulting HIP lit up the T cells.

Delong then worked backwards, using mass spectrometry to characterize the chemical architecture of his creation and confirming that it indeed occurs in mouse cells. With collaborators in Massachusetts and Australia, he found that HIPs could also be targets for human T cells isolated from the pancreases of deceased donors with Type 1 diabetes. Ten years of work and many hundreds of thousands of dollars from the National Institutes of Health, the American Diabetes Association, and the Juvenile Diabetes Research Foundation had led to what Haskins described as “a game-changer.”

Toward a cure

Delong, Haskins and colleagues are now working on developing therapies based on the HIP antigen that can turn off Type 1 diabetes in mice and, if it works, humans. With the help of the CU Technology Transfer Office, they filed a patent covering diagnostics and therapeutics based on HIPs on March 4, Haskins said.

But the work on Type 1 diabetes may be only the beginning. “This is a whole new type of antigen that could be related to other autoimmune diseases,” Delong said.

The team is particularly interested in whether hybrid peptides could be antigens in autoimmune thyroiditis, rheumatoid arthritis, multiple sclerosis, and Sjogren’s syndrome, he said. If you know the trigger, you can develop diagnostics to detect disease-driving T cells and, they hope, create therapeutics to prevent full-blown autoimmune disease from developing – for example, by using other hybrid peptides to re-educate T cells so they lay off the insulin-producing pancreatic beta cells.

The beauty of such antigen-specific therapy is its narrow targeting – potentially another application of personalized medicine, Haskins said. You take out only the T cells that are causing the problem, and not helpful T cells, too, as is the case with broad-spectrum immunosuppressants.

For Delong, his own diabetes played the role of antigen, attracting him to the field he’s now disrupted like a T cell, though in his case in a strictly positive sense.

“I probably would not be doing diabetes research if I didn’t have it,” he said. “But that doesn’t change anything about how much I enjoy doing diabetes research.”