Recipe for reprogramming

The science of induced pluripotent stem cells advances rapidly in mice

Here’s the recipe:

Take cells from an adult mouse. Using retroviruses, add four genes to the genome. Cultivate carefully for a few weeks.

Congratulations! You have produced induced pluripotent stem (IPS) cells. And these cells seem to be functionally equivalent to embryonic stem cells, able to differentiate into almost any body cell, without requiring an egg.

In the top image, neurons derived from IPS cells, transplanted into embryonic mice, matured in the brain and integrated with surrounding host neurons. The middle image is a closeup of the white box on the left, while the bottom image zooms in on a single neuron. Images: Courtesy of PNAS

Of course, this recipe vastly oversimplifies the extremely complicated biochemistry and weeks of highly expert work required to make IPS cells and keep them healthy.

But the core technique of reprogramming cells is now well understood in labs around the world. It’s only two years since such cells were first reported, but they apparently offer all the potential of embryonic stem cells without the legal, technical and supply problems.

Research on IPS cells is moving ahead at very high speed, and Whitehead Member Rudolf Jaenisch’s lab is at the forefront. In December 2007, researchers reported successful treatment of mice with a human sickle-cell anemia disease trait—the first proof-of-principle of therapeutic applications.

In April, scientists in the Jaenisch lab reported two more accomplishments with IPS cells: integrating neurons derived from IPS cells into mouse and rat brains, and creating IPS cells by reprogramming fully mature B cells.

Treating Parkinson’s disease
Neurons derived from IPS cells have been successfully integrated into fetal mouse brains and reduced symptoms in a rat model of Parkinson’s disease. “This is the first demonstration that reprogrammed cells can integrate into the neural system or positively affect neurodegenerative disease,” says Marius Wernig, a postdoc in the Jaenisch lab and lead author of the article, published in the Proceedings of the National Academy of Sciences.

Wernig and his colleagues created IPS cells by reprogramming adult skin cells using retroviruses to insert four genes (Oct4, Sox2, c-Myc and Klf4) into the cells’ DNA. The IPS cells were then differentiated into neural precursor cells and dopamine-generating neurons via techniques originally developed in embryonic stem cells.

In one experiment, Wernig transplanted the neural precursor cells into brain cavities of mouse embryos. The mice were naturally delivered and analyzed nine weeks after the transplantation. Transplanted cells formed clusters where they had been injected and then migrated extensively into the surrounding brain tissues. Using electrophysiological studies conducted by Martha Constantine-Paton from MIT’s McGovern Institute for Brain Research as well as structural analysis, Wernig also saw that the neural precursor cells that migrated had differentiated into several subtypes of neural cells, including neurons and glia, and functionally integrated into the brain.

Back from B cells Fully mature mouse B cells can be reprogrammed to IPS cells by adding one more gene in another round of the reprogramming process. The accomplishment highlights the power of the IPS cell approach and points toward mouse models that will aid in understanding autoimmune diseases such as multiple sclerosis and type 1 diabetes.

Illustration: Tom DiCesare

To investigate the therapeutic potential of the reprogrammed cells, Werning collaborated with Ole Isacson’s group at McLean Hospital/Harvard Medical School. The scientists used a rat model for Parkinson’s disease, a human condition caused by insufficient levels of the hormone dopamine in a specific part of the midbrain. To mimic this state, the dopamine-producing neurons were killed on one side of the rat brains. The researchers then grafted differentiated dopamine neurons into the rat brains.

Four weeks after surgery, the rats were tested for dopamine-related behavior. In response to amphetamine injections, Parkinson’s-model rats typically walk in circles toward the side with less dopamine activity in the brain. Eight of nine rats that received the dopamine neuron transplants showed markedly less or even no circling. Eight weeks after transplantation, Wernig could see that the dopamine neurons had extended into the surrounding brain.

“This experiment shows that in vitro reprogrammed cells can in principle be used to treat Parkinson’s disease,” says Jaenisch.

He and Wernig are optimistic that this work eventually could be applied to human patients, but caution that major hurdles must be addressed first. One is to find alternatives to the potentially cancer-causing retroviruses that transform the skin cells into IPS cells; researchers in the Jaenisch lab and many others are eagerly examining candidate methods. A second hurdle is to figure where and how to transplant the neural precursor cells with best results in humans.

Making IPS cells from B cells
Fully mature, differentiated mouse B cells can be reprogrammed to IPS cells without the use of an egg, according to a study Jaenisch lab investigators published in Cell. Scientists plan to use this technique to create mouse models that will aid in understanding autoimmune diseases such as multiple sclerosis and type 1 diabetes.

In previous research, IPS cells have been created from fibroblasts, a specific type of skin cells that may differentiate into other types of skin cells. Because there is no way to tell if the fibroblasts were fully differentiated, the cells used in earlier experiments may have been less differentiated and therefore easier to convert to the embryonic-stem-cell-like state of IPS cells.

B cells are immune cells that can bind to specific antigens, such as proteins from bacteria, viruses or micro-organisms. Unlike fibroblasts, mature B cells have a specific part of their DNA cut out as a final maturation step. “Once that piece of DNA is cut out, it can’t come back,” says Jacob Hanna, first author on the paper and a postdoctoral fellow in the Jaenisch lab. “Checking the genome gives us a way to make sure the resulting IPS cells were not from immature cells.”

Hanna and his colleagues began their experiment by generating IPS cells from immature B cells. As researchers had done in earlier work creating IPS cells from fibroblast cells, Hanna successfully reprogrammed the immature B cells into IPS cells by using retroviruses to transfer four genes (Oct4, Sox2, c-Myc and Klf4) into the cells’ DNA.

However, an additional gene, C/EBP-alpha, was needed to nudge mature B cells to become IPS cells.

Like IPS cells from earlier fibroblast studies, the IPS cells from both the mature and immature B cells could be used to create mice. The mice grown from the reprogrammed mature B cells were missing the same part of their DNA as the mature B cells, demonstrating that Hanna and his colleagues had successfully reprogrammed fully differentiated cells.

In addition to demonstrating the power of reprogramming, this work offers the promise of powerful new mouse models for autoimmune diseases such as multiple sclerosis and type 1 diabetes, in which the body attacks certain types of its own cells. For example, mature B or T cells specific for nerve cells called glia could be reprogrammed to IPS cells and then used to create mice with an entire immune system that is primed to attack only the glia cells, thereby generating a mouse model for studying multiple sclerosis.

Eventually, researchers will be able to study diseases by following a similar process with human cells, Jaenisch predicts. “In principle, this will allow you to transfer a complex genetic human disease into a Petri dish and study it,” he says. “That could be the first step to analyze the disease and to define a therapy.”