Peering through a microscope in Madison, Wisconsin, I watched my heart cells beat in a petri dish. Looking like glowing red shrimp without tails, they pulsated and moved very slowly toward one another. Left for several hours, I was told, these cardiomyocytes would coalesce into blobs trying to form a heart. Flanking me were scientists who had conducted experiments that they hoped would reveal whether my heart cells are healthy, whether they’re unusually sensitive to drugs, and whether they get overly stressed when I’m bounding up a flight of stairs.
It was snowing outside the office-park windows of Cellular Dynamics International (CDI), where I was observing an intimate demonstration of how stem-cell technologies may one day combine with personal genomics and personal medicine. I was the first journalist to undergo experiments designed to see if the four-year-old process that creates induced pluripotent stem (iPS) cells can yield insight into the functioning and fate of a healthy individual’s heart cells. Similar tests could be run on lab-grown brain and liver cells, or eventually on any of the more than 200 cell types found in humans. “This is the next step in personalized medicine: being able to test drugs and other factors on different cell types,” said Chris Parker, CDI’s chief commercial officer, looking over my shoulder.
CDI scientists created the little piece of my heart by taking cells from my blood and reprogramming them so that they reverted to a pluripotent state, which means they are able to grow into any cell type in the body. The science that makes this possible comes from the lab of CDI cofounder and stem-cell pioneer James Thomson of the University of Wisconsin, the leader of one of two teams that discovered the iPS-cell process in 2007. (The other effort was led by Shinya Yamanaka of Kyoto University.) The results are similar to the special cells that appear in embryos a few days after fertilization.
Since late 2008, the company has been manufacturing cardiomyocytes and mailing the frozen cells on dry ice to academic scientists to study how these cells work, and to researchers in the pharmaceutical industry to use in early tests of drug candidates. One important reason to use the cells is that they could reveal whether drugs are toxic to the heart, information that other types of testing can miss. “Several drugs have made it to the market that have cardiotoxic profiles, and that’s unacceptable,” Parker says. He says that the cardiomyocytes derived from iPS cells are a huge improvement over the cadaver cells sometimes used to test potential drug compounds. Unlike the cadaver cells, IPS-generated cells beat realistically and can be supplied in large quantities on demand. What’s more, iPS-generated cells can have the same genetic makeup as the patients they came from, which is a huge advantage in tailoring drugs and treatments to individuals. These made-to-order cells are not cheap, however. Cellular Dynamics’ CEO, Robert Palay, says they cost about $1,500 for a standard vial of 1.5 million cells.
An especially sensational prospect is that iPS cells could be transplanted into patients so they could regenerate diseased or damaged spines, brains, hearts, or other tissue—a proposition that is particularly enticing because these cells wouldn’t be rejected by the host’s body. They could also defuse the political controversy around embryonic stem cells, because they may one day make it possible to harvest pluripotent cells without destroying a human embryo.
Transplantation, however, is years away for most tissue types, says Alexander Meissner, a Harvard University stem-cell researcher. “It’s not trivial to regenerate brain tissue,” he says. “This is going to take longer than people think.” Thomson agrees. “Talk about transplantation has been a kind of irrational exuberance,” he says. The process of using iPS cells to create new tissue still poses certain dangers: some cell lines, for example, harbor mutations that could lead to cancer, and in some cases cells retain a faint chemical memory of their previous identity as skin or blood cells.
Thomson believes these are temporary setbacks. “We have had bone marrow transplantation for a long time, which is essentially stem cells,” he says. “And work is being done right now on using iPS cells to repair macular degeneration. But repairing damage to the nerves in a spine is much more difficult.” Others share his cautious optimism. “Virtually everything about iPS cells is overhyped,” says Chris Austin, director of the Chemical Genomics Center at the National Institutes of Health. “But for the purpose of testing drug candidates, I think the possibilities are considerable, and we and lots of other people are pursuing this. There are lots of problems. Are iPS cells really normal? How do you get enough pure differentiated cells? But the potential is definitely there.”
Sticking to Science
I first visited James Thomson on another snowy, frigid day in Wisconsin in 2008, a few weeks after the publication of his paper announcing iPS cells derived from human cells. A scrappy, no-nonsense man in a casual sweater and beat-up Dockers, he sat in a small office adorned with tropical fish, ferns, and an antique dartboard and discussed his original discovery of human embryonic stem cells in 1998. His work set off a storm of protest: opponents argued that destroying a human embryo to harvest its stem cells is tantamount to murder. President George W. Bush restricted most federal funding for embryonic-stem-cell research in 2001, and critics have continued to vilify Thomson, although he tries to keep a low profile. “I don’t talk much about it,” he said. “I stick to the science.”
The creation of iPS cells in 2007 seemed like an elegant bookend to the 1998 finding, because it offered a new way to produce stem cells that can differentiate into any cell type—one that might actually be better, because the cells would be genetically identical to patients’ own. “It was a relief that we might have a solution to this political and ethical situation,” Thomson said. The breakthrough, however, was a surprise. “We knew that the iPS process was a possibility,” he said, “but when we started out, I was sure it would take 10 years at least.” Thomson and a Wisconsin postdoc, Junying Yu, set out to create iPS cells by modifying skin cells with “regulator” genes normally found only in embryos. The method, he said, “surprised everyone by working.”
Thomson cofounded CDI in 2007, around the same time that several other stem-cell luminaries became involved in iPS-cell companies. These would-be competitors, however, are primarily focused on creating therapeutics. They use iPS cells to help identify and develop drug candidates and to design processes that might one day lead to transplantation. So far CDI has no serious competitors in the market to sell iPS-generated cells in volume for use in research and drug screening. In part, this is because Thomson and his scientific team have been working longer to overcome difficulties in industrializing the technology. “Making iPS cells that are functional in large quantities is tough,” says Harvard’s Meissner.
Privately held, the company has not detailed its performance, but its CEO told a local newspaper that CDI gets “multimillions” in revenues from selling its heart cells to about 40 customers, including most large pharmaceutical companies. Next year the company plans to roll out iPS-generated liver, brain, and blood cells.
“This is a game-changer,” says stem-cell biologist Sandra Engle, a senior principal scientist at Pfizer who has used CDI’s cells. “Before CDI, these cells were very difficult to obtain, and we would only get tiny amounts. This doesn’t work for high-throughput testing for drugs.” For Kyle Kolaja, global head of predictive toxicology screens and emerging technologies at Roche, the benefit of the CDI cells is that they behave like “real” cells. “They are already having a major impact on drug safety and development,” he says. “They have already changed what we’re doing.”
Although companies like Roche and Pfizer are currently using iPS cells simply to screen potential therapeutics for toxicity and other characteristics, someday iPS-based tests could be performed on individual patients to see whether they are at particular risk for side effects. Euan Ashley, a cardiologist at Stanford University, is trying to use iPS cells to help diagnose and treat a 16-year-old boy with early symptoms of dilated cardiomyopathy, a potentially fatal disease in which the heart swells and weakens. “This is the sort of severe genetic disease that runs through families that we think can benefit from iPS technologies and genomics,” says Ashley. He has scrutinized the boy’s DNA for telltale genetic markers associated with the disease and has tested his brother and parents to see if they carry the markers as well. The Stanford team plans to create iPS cells by reprogramming skin cells taken from the family and then induce them to differentiate into cardiomyocytes bearing the characteristic genetic variations. By studying the biochemistry of these heart cells, the scientists hope to gain clues to how they might respond to various drug candidates.
“We will use the iPS cells to check the differences between this child and others with and without the condition,” says Ashley, “and to test what drugs will work best for the boy and other impacted family members.” Ashley says one goal is to develop tests to determine how the genetic variations actually affect the cells. “The importance of genetic factors will be reflected in these cells,” he says.
Other clinicians and labs are also using iPS cells in experiments intended to shed light on disease. For instance, researchers at the Salk Institute are studying iPS-derived neurons from people with schizophrenia to see how they differ from normal neurons, and they will examine what happens when the cells are exposed to antipsychotic drugs. At the NIH, a group is studying iPS-generated cells from patients with a fatal genetic disorder known as Niemann-Pick disease type C. Other researchers have proposed using iPS-generated cells to test the effects of toxic chemicals such as mercury and pesticides.
The hope, say researchers, is to create a library of iPS cell lines from people who have specific symptoms or behaviors associated with a particular disease. Roche has started a program with Massachusetts General Hospital in Boston to create cell lines that reflect different types of heart disease; the results could help the company develop drugs. This summer, CDI and the Medical College of Wisconsin announced a $6.3 million grant from the NIH to create iPS-generated heart cells from 250 patients who have left ventricular hypertrophy, a condition that causes high blood pressure and increases the risk for cardiovascular disease.
Scientists are still a long way from using iPS cells routinely to diagnose disease or offer individual prognoses. The NIH’s Austin cautions that individual cells tell only part of the story of what happens in the dynamic system that is the human body. “In some cases, you don’t have a cell that can give you a real answer about a disease like depression,” he says. “What cell type do you use for that?”
I launched my own iPS journey in a small Quest Diagnostics clinic on a leafy street in San Francisco. Wrapping a rubber tube around my arm, the phlebotomist stuck in a needle to withdraw several vials of blood that would be shipped on ice to Madison. Once they got to CDI, technicians cracked open my white blood cells and used a bioengineered retrovirus to introduce “master transcription” genes into their DNA. These genes reprogrammed the cells so that when they replicated, the results were pluripotent cells rather than more white blood cells. Their transformation into functioning iPS cells took several months of coaxing, purification, and verification that cost about $15,000, which the company paid on my behalf. Once my pluripotent cell line was humming along, the scientists at CDI tweaked a few cells to make them differentiate into three types of heart cells—which I first saw pulsing in a video clip they e-mailed to me.
In Madison, nearly a year after giving up my blood, I was just a bit anxious as I stared at my beating heart cells. I was about to get a rundown on the experiments CDI had performed to demonstrate what these little bundles of bioengineered cytoplasm and nuclei might say about my health and my sensitivity to various drugs.
Chris Parker and the company’s product manager for cardiomyocytes, Blake Anson, took the lead in walking me through a series of assessments that began with tests “to make sure these cells are still you,” said Parker. They showed me a slide of the 23 paired chromosomes taken from my original blood sample and compared it to a slide showing the chromosomes taken from the cardiomyocytes. They had also run a simple genetic comparison using 16 DNA markers, a test used by law enforcement that provides a quick, relatively cheap way to assess whether two samples match up. When my manufactured cells passed muster, the scientists moved to step two: seeing if they behaved like real cardiomyocytes.
First they buzzed the cells with electricity to check the range in duration of the action potentials—the electrical impulses that drive cardiac contractions. Then they measured the beats of the cells in the aggregate against a kind of EKG waveform like those that appear as up-and-down pulses on a hospital monitor. My cells appeared normal.
A third test pitted the cells against two drugs. One was epinephrine, which triggers the fight-or-flight response and speeds up a person’s pulse. “We can see this here: beat, beat, beat,” said Parker, showing me a slide with an EKG line. “Your heart rate increases dramatically, so that means you’re okay—you can run from that bear.” The scientists then dropped in a “sympathetic agonist,” a drug that slows the heart way down. “So your cells can relax after running from that bear,” said Parker. When I sent Euan Ashley my test results, he verified my persistent normalcy—and confirmed that the cells in question were what they were supposed to be. “These tests prove that the cells are cardiomyocytes,” he said, “which at this early stage in this science is important.”
A few weeks later, CDI ran another round of experiments that subjected my cells to drugs with known toxic side effects. First came Hismanal, an antihistamine, and Propulsid, a drug to treat gastrointestinal distress. Both medications were pulled from the market in many countries, including the United States, because they were associated with rare but potentially life-threatening heart arrhythmias. “This propensity is due to the unanticipated and unwanted side effect of both drugs blocking and disrupting the normal activity of a specific ion channel in the heart,” said a report e-mailed to me from CDI. “Both drugs had similar effects on David Duncan’s iPS-derived cardiomyocytes: a dose-dependent increase in the duration of the action potential … Prolonged action potential durations are a recognized trigger for cardiac arrhythmia that can result in sudden death.”
For a second round of pharmaceutical testing, the scientists exposed my cells to two cancer drugs: Gleevec, used mostly to treat some forms of leukemia, and Sutent, used to treat tumors in the stomach, bowel, and esophagus. Both drugs have side effects that include damage to the heart, though they remain in use because the diseases they treat are so serious. “In vitro tests on David Duncan’s iPS-derived cardiomyocytes demonstrated that both drugs had adverse effects,” said my report, “and that the Gleevec-mediated effect may have been caused by disrupting mitochondrial function.” Again, the reactions of my cells were not atypical, although the researchers told me that if I had cancer, further testing might turn up specific responses that could help a physician decide which medications were best for me.
Ashley told me that iPS-generated heart cells offer great potential as a way to test cancer treatments. “Chemo drugs are really hard on hearts, and on heart cells,” he said. “If this technology can help, that will be really important.”
CDI has told me that as the science unfolds, it may run tests based on the extensive DNA sequencing I had done for a recent book, Experimental Man. I’d be especially interested in a test that could determine how worried I should be about a genetic risk factor for side effects of cholesterol-lowering statins. According to my genetic profile, I have a substantial risk of myopathy—muscle weakness—if I take certain forms of these drugs. However, this condition is due to a malfunctioning enzyme produced by the liver, not the heart, so finding out depends on whether CDI is willing to create liver cells from my iPS line.
Before I left the CDI lab, I took one more look at my heart cells pounding away in their petri dish in a sort of freakish mambo, and I wondered when such banks of individual cells would become a routine part of medical care. Many obstacles remain before this can happen, including the high cost of making the cells. Yet despite the expense, says Thomson, “there will be people that will want to do this—wealthy early adopters who want to know about a disease or a drug. Or some people might do it because they think having their beating heart cells is cool.”
As for me, I’m still amazed that the cardiomyocytes in the dish are part of me—let alone that they might one day be used as stunt doubles for my real cells.
David Ewing Duncan is a San Francisco-based writer. His most recent book is Experimental Man: What One Man’s Body Reveals about His Future, Your Health, and Our Toxic World.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.