Originally published 2017-07-24.

In April 2017, researchers at UT Health San Antonio announced that a gene-transfer treatment had reversed diabetes in mice for a full year. The idea was to teach other cells in the pancreas to make insulin, replacing the work of the beta cells that diabetes destroys or wears out. It was a promising animal result, not a human cure: the therapy had not been tested in people, and the team said larger animal studies had to come first.
Here is what the scientists actually did, why the approach drew attention, and how to read headlines that say diabetes has been "cured in mice."
What the San Antonio team reported
The work came from Bruno Doiron, PhD, an assistant professor of medicine, and Ralph DeFronzo, MD, professor of medicine and chief of the Division of Diabetes at what was then called the University of Texas Health Science Center at San Antonio. Both are named as inventors on a US patent granted in January 2017, and the university said it was spinning out a company to develop the treatment commercially. The study itself appeared in the journal Current Pharmaceutical Biotechnology.
In diabetic mice, the treatment brought blood sugar under control and kept it there for a year. Doiron summed up the result in the university's announcement: "We cured mice for one year without any side effects." In the same breath he stressed that a mouse model calls for caution.
Why beta cells matter in both types of diabetes
Insulin is the hormone that lets the body move sugar from the blood into cells. In healthy people, only one cell type makes it: the beta cells, clustered in small islands within the pancreas.
- Type 1 diabetes is an autoimmune disease. The immune system attacks and destroys beta cells, so the body makes little or no insulin. People with type 1 need insulin every day to survive.
- Type 2 diabetes develops when the body responds poorly to insulin (insulin resistance) and the beta cells gradually fail to keep up. Many people manage it with lifestyle changes and tablets, but some eventually need insulin injections too.
According to Doiron, symptoms of type 1 usually appear only after about 80 percent of beta cells are gone, and restoring roughly 20 percent of insulin-making capacity might be enough to free a patient from injections. That is the gap the gene-transfer approach aimed to fill.
How the gene-transfer approach works
The pancreas is not just beta cells. Most of it is made of cells that produce digestive enzymes and other substances. The San Antonio team's idea was to reprogramme some of these cells so they also release insulin.
To do this, they built lab-made DNA sequences and delivered a "cocktail" of three molecules into the mouse pancreas using a virus as the carrier, or vector. Viral vectors are standard tools in gene therapy: the virus is modified so it cannot cause disease, but it can still enter cells and deliver new genetic instructions. Once inside, the instructions prompted the cells to make insulin.
Two features made the result stand out. First, the reprogrammed cells released insulin only when glucose was present, much as natural beta cells do. Second, because these cells are not the target of the autoimmune attack in type 1 diabetes, the team expected the immune system to leave them alone. Long-term human data would be needed to confirm that.
In people, the researchers proposed reaching the pancreas by endoscopy, passing an instrument through the throat and placing the molecules into the organ through a catheter, rather than by open surgery.
Why low blood sugar is a big deal
Living with insulin means constant balancing. Too little and blood sugar climbs, damaging blood vessels, nerves, eyes and kidneys over time. Too much and blood sugar can crash, a state called hypoglycaemia, which can cause shaking, confusion, seizures and, in severe cases, loss of consciousness.
Insulin injections and some diabetes tablets carry this risk because they keep working whether or not blood sugar is high. A treatment that releases insulin only in response to glucose would, in theory, regulate itself second by second. That is why Doiron described hypoglycaemia as a major problem in type 1 care and highlighted the glucose-responsive behaviour of the altered cells as the most remarkable part of the work.
The picture for type 2 was more limited. Texas Public Radio reported that people with type 2 would likely still need a daily pill to address insulin resistance, since extra insulin does not fix the body's poor response to it.
Checking the claims: mice, side effects and FDA approvals
"Cured" applies to mice. Mouse models of diabetes are useful but imperfect; many treatments that work in rodents never succeed in people. DeFronzo put it plainly to Texas Public Radio: "I'm the first person to say mice are not men."
"No side effects" means none were observed. A lack of visible problems in a small group of mice over a year does not show that a therapy is free of risk. Gene therapies can raise concerns about immune reactions to the viral vector, where the new genes end up, and long-term effects, all of which need careful study.
The FDA approval figure needs context. In the original announcement, DeFronzo said viral gene transfer had been approved by the FDA nearly 50 times for various diseases. That figure is hard to square with the record: when it was made in April 2017, the FDA had not yet approved any gene therapy for sale in the US. Its first approvals came later that year, including Luxturna for an inherited form of blindness in December 2017. The comment may have referred to authorised clinical studies rather than marketed treatments. Either way, those approvals concerned other diseases, not this diabetes therapy.
The road from a mouse study to a treatment
The team laid out the next steps in 2017. Before testing in people, they needed studies in larger animals with an endocrine system closer to ours, such as dogs, pigs or monkeys. That stage was estimated at around $5 million, with Texas Public Radio citing a range of $5 million to $10 million. Positive results would support an Investigational New Drug application to the FDA, the permission needed to start human trials. The researchers hoped to reach that point within about three years.
Timelines like this often slip, because funding, manufacturing and safety testing all take longer than expected. If you or someone you love lives with diabetes, a few habits help you read research news well:
- Check whether the study was in cells, animals or people.
- Look for a registered human trial on ClinicalTrials.gov before treating a result as near.
- Be wary of clinics or products claiming to sell a diabetes "cure."
- Keep taking your prescribed treatment, and talk to a doctor about any new therapy you read about.
The San Antonio work belongs to a wider effort to restore the body's own insulin production, alongside islet transplants and stem-cell-derived beta cells. Whatever its final fate, it showed a creative way of thinking about the problem: rather than replacing lost beta cells, persuade their neighbours to take over the job.
Sources: UT Health San Antonio, 20 April 2017: Diabetes breakthrough increases insulin-producing cells; Texas Public Radio, 19 April 2017: No insulin shots? Diabetes 'cure' under study in San Antonio; Wikipedia: Gene therapy; first reported via Medical Xpress, provided by the University of Texas Health Science Center at San Antonio.

COMMENTS