A fifth group turns skin to neurons, creating a model for Alzheimer's disease

A few weeks ago, my colleague used this space to discuss the second and third papers showing teams had turned skin cells directly into neurons, noting that this replication of research results is essential to verifying the initial breakthrough while refining and improving it. She noted that only after much replication and refinement would she or anyone else want the resulting cells for therapy. Since then a fourth team reported another technique and now a fifth group is reporting the more likely short term benefit—a disease in a dish model for a neurodegenerative disease.

The first two papers came from work in the CIRM funded facility at Stanford University. The breakthrough paper in May from Marius Wernig reported a slow and inefficient process for using certain factors to directly reprogram skin cells into nerve cells without first taking them through an embryonic-like state. The follow-up paper in late July came from Stanford colleague Gerald Crabtree and showed marked improvement in efficiency and the nerve-like functioning of the cells. That same week at team from Milan reported another efficient system for creating nerve cells but this time directing them to become dopamine-producing cells like those lost in Parkinson's disease. And last week at team at the Gladstone Institutes in San Francisco, using another CIRM funded facility, reported yet another technic that also improved efficiency.

Now, a team at Columbia reported in today's issue of the journal Cell that they had developed a fifth way to accomplish this direct conversion of skin to nerve, and had done so with both skin from normal subjects and skin from patients with the inherited form of Alzheimer's disease. In both cases the cells matured and behaved like neurons, responding to neurotransmitters by letting ions like sodium and potassium cross the cell membrane. However, the cells derived form patients were also clearly abnormal. They had altered ability to process and transport the amyloid precursor protein (APP) and a resulting increase in production of amyloid beta, which has long been a suspect in the disease, but depending on what year you look at the literature, it is theorized to be a culprit or an artifact. This disease in a dish model may help to answer that question.

DG

Improved technique for directly converting skin to neurons

This is the way things often go in science: One group announces a breakthrough. Yah! Then for the next several years, scientists all over the world replicate and improve on that breakthrough until it's finally believable and widely useful.

To people outside science who read about the initial breakthrough, this may look a lot like scientists twiddling their thumbs, sitting on new therapies. But really, do you want a therapy based on a breakthrough that may or may not be real? Right, neither do I.

A paper from Marius Wernig's lab at Stanford University is a great example of this process. In January, 2010, Wernig's lab had a paper in Nature announcing their transformation of mouse skin cells directly into neurons. This was exciting work, bringing with it the possibility of directly converting skin from a person with a neuronal disease into neurons that can be studied in the lab. But that work was in mice, and one thing we know from past research is that mice are most certainly not humans.

About a year and a half later, Wernig replicated his work with human cells in another Nature paper, but the transformation was much less efficient than it was with mouse cells (here's our blog entry on that work). It took weeks for the transformation to take place, only 2 to 4 percent of the skin cells transformed into neurons and those neuronal cells were on the wimpy side. It's still exciting work — I mean how cool is it that human skin can be turned into neurons with the addition of just four molecules. But ready for therapeutic prime time? I think not.

Now we've entered the next stage where scientists all over the world incrementally improve upon the original work until it's good enough, fast enough and efficient enough to be broadly useful. One such improvement came from the Stanford University lab of Gerald Crabtree, who published his findings in a Nature paper last week.

Crabtree's lab employed two of the four factors that had been effective for Wernig, but supplemented those with a different kind of molecule — called microRNAs. This change dramatically improved how efficiently the skin cells converted to neurons, and produced neurons with much stronger electrical signaling. Another group from Milan published a paper in early July using three different factors to coerce the transformation from skin to neuron. In their case, the neurons were more like those that are lost in Parkinson's disease, known as dopaminergic neurons.

A Stanford press release quotes Crabtree:

“It’s been a long time in coming to this,” said Crabtree. “But science often progresses in leaps and starts, and then all of a sudden many scientists come to the same position at the same time. Now these studies have come out, and more will be coming, all of which are going to say that not only can you can make neurons different ways, but also you can make neurons of different types.”

At this point it's too soon to know which, if any, of these techniques is going to become most widely used. We can probably expect to see more improvements on these approaches coming out of some labs, while other labs start figuring out how this revolutionary transformation can be used to treat or understand disease. Crabtree's lab, for example, says they are already taking skin cells from people with Down's syndrome and transforming those into neurons in order to understand the disease and look for therapies.

A.A.