Bungled Insulin Production May Be A Culprit In Diabetes, University Of Michigan Researchers Say

But that misfolding occurs far more often in the cells of mice that must overproduce insulin, as people with type 2 diabetes do, than those without, the researchers from the University of Michigan and Pennsylvania State University report.

And, they find, the incorrectly folded forms of insulin can't leave their production site inside the cell nearly as easily as the correct form. The resulting buildup of misfolded molecules inside the cell may also be enough to stress its cleanup mechanisms — and may even contribute to the cell's death.

The discovery, published early online by the Journal of Biological Chemistry, may help explain why people with all forms of diabetes eventually make less insulin and experience the death of insulin-producing cells in the pancreas. The destruction of pancreatic beta calls is the hallmark of both type 1 (juvenile) and late-stage type 2 (adult) diabetes.

“Clearly, the occasional misfolding of the insulin precursor molecule doesn't faze the cell. But abundant production of misfolded forms, and the resulting effect on beta cells, may be an important factor in diabetes,” says senior author Peter Arvan, M.D., Ph.D., the Brehm Professor and Chief of the Metabolism, Endocrinology & Diabetes Division at the U-M Medical School.

Adds lead author and U-M research associate Ming Liu, M.D., Ph.D., “It's an exciting first step, but more research is needed to better understand this effect.”

The new U-M research focuses on proinsulin, the molecule made in the endoplasmic reticulum area of the cell that later gets chopped to make insulin. Just seconds after proinsulin is made, the long molecule folds like origami, as six sulfur atoms within the molecule pair up to form three disulfide bonds. The locations of those sulfur atoms, and the bonds, have remained the same throughout the evolution of mammals, likely because of insulin's importance in the body's metabolic system.

For the first time ever, the U-M researchers were able to show that normal rat and human cells produce misfolded, as well as normally folded, proinsulin molecules. They made the discovery by separating proinsulin from cells, and analyzing the molecules using a special technique that allowed them to see when the disulfide pairs were mismatched.

They also showed far higher levels of misfolded proinsulin in mice with gene mutations that made them prone to diabetes, and in mutants of proinsulin itself. Some of the mutants made no normally-folded insulin at all.

Previously, scientists have suspected — but have been unable to show — that some proinsulin is misfolded in normal, and diabetic, beta cells. The U-M and Penn State team was able to demonstrate the effect by using a form of a technique called electrophoresis that let them separate the different kinds of proinsulin without denaturing the disulfide bonds. The specific technique they used is called non-reducing tris-tricine-urea-SDS-PAGE.

The team examined proinsulin production in pancreatic islet cells from normal rats, from cells that made proinsulin using a human gene, and from mice that are prone to diabetes.

One mouse strain, called Akita, has a gene mutation that leads to the production of proinsulin that lacks a key amino acid and interferes with the function of any normal insulin made in the cell; this is called a dominant-negative diabetes and is similar to a rare form of type 2 diabetes that occurs in young adults. Because this strain of rat creates little or no functioning insulin, it can be used as a model for human type 1 diabetes, in which the pancreas stops producing insulin. Another diabetic mouse had been bred to lack a gene called PERK, which limits the number of beta cells in the pancreas, leading to diabetes.

In the normal and transgenic cells, the electrophoresis results showed that three kinds of proinsulin were made — a large quantity of normal molecules, and smaller quantities of two “isomers” or variations. When the researchers looked at the secretion, or the ability of the proinsulin to leave the endoplasmic reticulum and be prepared for entry into the bloodstream, they found that the normal proinsulin was almost always secreted — but less than half of the two isomer forms was secreted.

The researchers also created mutations in the proinsulin gene beyond the one that causes the defect in Akita mice. When the mutation affected one of the less-important disulfide bonds, the result was the production of a small quantity of proinsulin that could still be secreted, and a large quantity of proinsulin that was secreted less than half the time. But when the mutation interfered with the most important disulfide bond, none of the resulting proinsulin was secreted. The researchers also found evidence that some of the misfolded proinsulin was “cleaned up” by the cells; when they interfered with the cleanup mechanism, the molecules were still detectable hours later.

In the PERK-knockout mice, which over-produce insulin for the first few weeks of life before starting to lose their beta cells, the situation was different. When compared with normal mice, the PERK-knockout mice made far more proinsulin, especially when their blood glucose was kept high — and under those conditions, more proinsulin was misfolded than in other mice.

Although the new research didn't directly show that the misfolded proinsulin molecules are toxic to beta cells, Arvan says, the results are consistent with that theory. “We've made the first step of showing that these misfolded forms are made under normal conditions, and in higher abundance under diabetic condition,” he says. “The next step is to see how that affects cell function.”

Arvan and his team are continuing their work with help from a major gift by Bill and Delores Brehm, of McLean, Va. The Brehms gave $44 million to U-M in November to support research and facilities aimed at finding a cure for type 1diabetes, which Dee Brehm has had for more than 50 years.

The current research was funded by the National Institutes of Health and the American Diabetes Association. In addition to Arvan and Liu, the authors include Yulin Li and Douglas Cavener of the Department of Biology at the Pennsylvania State University.

Reference: J. Biol. Chem., Papers in Press, online February 2005; 10.1074/jbc.C400475200