Over the last decade, scientists have focused on stem cells as a solution because of their natural ability to grow into new tissues that can be transplanted into patients.
Of particular interest were embryonic stem cells, which can grow into any type of cell in the body with the proper chemical prodding -- though figuring out those regimens has not always been easy. The research was also controversial because the cells had to be harvested from days-old embryos, which were destroyed in the process.
In recent months, researchers have embraced a new technique that involves rewinding adult cells to an embryonic state by turning on a set of four genes that are active during early development.
The Harvard researchers wondered if they could find a shortcut.
"We just asked, sort of like an undergraduate, the simple question, 'Why should you have to go all the way back to the beginning? Could you go directly from one cell type to another?' " Melton said.
They chose to study diabetes, a disease that has been a central focus of stem cell research.
Patients with Type 1 diabetes need new beta cells to make insulin because their original beta cells have been destroyed by their immune systems.
Insulin is crucial for metabolizing sugar, and without it, patients must monitor their blood-sugar levels every few hours and inject themselves with insulin up to five times a day.
Melton said he is obsessed with finding a way to treat patients with Type 1 diabetes, including his own children Emma and Sam.
"I wake up every day thinking about how to make beta cells," he said.
Once he decided to try direct reprogramming, his team identified nine key genes that are active in mature beta cells and their close relatives.
They started turning them on and off using specialized proteins, known as transcription factors, that bind to specific parts of DNA. Every possible combination was tried to determine which were necessary to make insulin-producing cells. The researchers ultimately determined that only three were essential to the process, and they were activated by the proteins Ngn3, Pdx1 and Mafa.
The researchers injected mice with a virus that specifically infects pancreatic exocrine cells, the type that make up about 95% of the pancreas. The proteins carried by the virus turned on the dormant genes.
