Move over mice. Human induced pluripotent stem (iPS) cells are making strides to become the next best thing in translational research—disease-specific human cells grown in a dish. Using a variety of approaches, researchers have generated stem cells from mature adult cells of disease-afflicted patients and subsequently differentiated them into the various tissue types involved in the disease.
“The idea is that you can have a pluripotent stem cell line from a patient that already contains all the genetic background of the disease,” says Gustavo Mostoslavsky, a stem cell researcher at the Boston University School of Medicine. Now that the generation of iPS cells is “routine,” he adds, scientists can use the method to generate in vitro disease models, from which they can learn about molecular causes, as well as potential preventions and treatments.
The strategy is proving particularly valuable for neurodegenerative diseases, in which it is not easy to safely and ethically extract affected cells of the brain. Instead, researchers can remove more accessible cells, such as those of the skin, regress them into a pluripotent state, and then re-differentiate them into neurons. Furthermore, iPS cells can be expanded in culture and/or frozen for years, providing an unlimited supply of cells from a single patient that can be used to create any cell types needed for the study of a particular disease, now or in the future.
iPS cells allow scientists to strip a disease down to the “pure essence of what a genetic background is doing to alter function,” says Chad Cowan, an iPSC researcher at the Harvard Stem Cell Institute.
In July 2008, Kevin Eggan and colleagues at Harvard generated motor neurons from the skin cells of an 82-year-old woman with amyotrophic lateral sclerosis (ALS)—the first successful generation and subsequent differentiation of iPS cells from a patient with a chronic disease. The event was featured as TIME magazine’s medical breakthrough of the year. But the experiment did not address whether iPS cells retain the characteristics of ALS in a dish—a critical element of disease modeling.
Many early iPSC studies actually failed to identify any characteristics or traits of a disease in patient-derived cells. Then, in December 2008, Allison Ebert and colleagues at the University of Wisconsin-Madison published the first proof of concept that patient-derived iPS cells exhibited a defect associated with a disorder, in this case spinal muscular atrophy (SMA), a degenerative neuromuscular disease.
The team generated iPS cells from skin fibroblasts taken from a child with SMA, which is caused by mutations in the survival motor neuron 1 (SMN1) gene that reduce SMN protein expression. The child’s iPS-derived motor neurons demonstrated this reduced SMN expression and showed selective motor neuron death, another characteristic of the disease. Neurons derived from child's unaffected mother, with an otherwise similar genetic background, did not. The child’s cells also responded to compounds known to increase SMN protein in patient fibroblasts. “Because this is a human motor neuron culture system, we can study the natural disease” directly in afflicted human cells, says Ebert, which is likely more relevant to the human condition than animal models.
Since Ebert’s success, more advanced methods of iPSC generation, better understanding of the limitations of iPSC technology, and the addition of compounds to stress differentiated cells into producing a disease phenotype are yielding more and more disease-specific iPSC models.
In 2010, for example, a team led by Alysson Muotri at the Salk Institute for Biological Studies created an iPSC model for Rett syndrome, a genetic neurodevelopmental disorder, from Rett patient fibroblasts. The iPSC-derived neurons displayed fewer synapses, smaller cell bodies, altered calcium signaling and electrophysiological defects. This year, Muotri, now at the University of California, San Diego, also used iPS cells to create a model of an inherited form of ALS, which similarly demonstrated characteristics of the disease, including reduced levels of a key ALS-associated protein. And in March, Stanford University researchers did the same with fibroblasts from a patient with Parkinson’s disease, a degenerative disorder characterized by progressive neuronal death.
Thus far, however, successful iPSC models are largely limited to neurological disorders. This is partly because scientists have concentrated their effort on differentiating iPS cells into neurons, since they are a precious commodity: scientists can biopsy breast cancer cells and study them in a dish, but they can’t remove neurons from an individual with Alzheimer’s.
But because some adult cells have limited lifespans and often only survive for a few days in culture, iPSC technology is useful for studying even the more accessible tissue types, such as liver or fat cells. iPS cells can provide researchers with “a renewable source of those [cells], that always have the same genetics again and again,” said Cowan. “[It] really lets us dissect what’s gone wrong in a particular individual to give rise to a syndrome.” Using iPS cells differentiated into adipocytes, for example, Cowan and his team have studied both rare metabolic disorders like lipodystrophy, in which individuals lack fat, and more common disorder like insulin resistance.
iPSC technology also holds promise for studying more complicated disorders, such as Crohn’s disease, an inflammatory bowel disease. Crohn’s is notoriously difficult to study, as it appears to have both genetic and environmental components and involves many bodily systems. With iPSC technology, Mostoslavsky and colleagues at BU have made iPS cells from Crohn’s patients that can then hopefully be differentiated into all the cell types implicated in the disease, including neurons, macrophages, epithelial cells, and more. Then, says Mostoslavsky, the cell types can be combined together in vitro to see how they interact with each other and environmental factors like pathogens.
“For first time, this is a great chance to try to model this disease in vitro, using all potential elements involved in the disease,” he says.
Soon, iPS cells may be even more malleable than before. In a study published last Thursday (July 14) in Cell, Rudolf Jaenisch and colleagues demonstrated a new technique to manipulate individual genes in iPS cells. The ability to site-specifically modify the genomes of iPS cells will allow scientists to create disease-specific cell lines and controls that are genetically identical, except for a single genetic change.
Still, as with other model systems, several challenges remain. In February, Joseph Ecker and colleagues at the Salk Institute demonstrated that iPS cells have "hotspots" in their genomes that are not completely reprogrammed, and these hotspots are retained even after the cells are differentiated into adult somatic cells. Scientists have also found that iPS cells may be more or less able to differentiate into certain tissues because of retained DNA methylation patterns, or an “epigenetic memory.”
Additional differences may stem from the techniques used to reprogram the cells (e.g., viral vs. protein-based, etc.), which can result in different cellular and molecular properties. Furthermore, it can be difficult to generate a purified population of either iPS cells or iPSC-derived cells, says Ebert. “The results we’re getting might not be totally accurate if it is a mixed population of cells.”