In labs round the world, scientists are trying to close the gap between promise of stem cells and real-world therapies. This is the story of one.

A Month in the Life of a Stem Cell Lab

Yi Yu, a research assistant in the Melton lab, holds flasks containing human embryonic stem cells. Researchers have developed a recipe to turn them into beta cells, which secrete insulin in the pancreas. In what they call the Foundry, they hope to produce 200 flasks of beta cells this year. (All images by Chloé Hecketsweiler)

Pluripotent stem cells have the potential to generate any human cell type and researchers have shown that they may be used to repair damaged tissues and organs in the body. But what looks good in the lab doesn’t always translate to the clinic. In laboratories around the world, thousands of scientists are trying to close the gap between promise and real-world therapies.

For a month and a half, starting in January, I was embedded in the daily life of one such group of scientists — Douglas A. Melton’s laboratory at the Harvard Stem Cell Institute. I watched their experiments, learned about the complex science of stem cells, and talked with the researchers about their work and hopes. I was allowed to take pictures, and for this photo essay I tried to pick out moments and details that I found revealing, although scientists may see them as business as usual.

Melton’s lab focuses on diabetes, a disease that affects almost one in 10 Americans. In the lifelong form of the disease (known as type 1), the body’s immune system attacks and destroys the beta cells that produce insulin in the pancreas. Diabetic patients must rely on daily injections of insulin to control the level of sugar in their blood.

Melton’s team has invented a protocol to turn embryonic stem cells into beta cells and has shown that they effectively secrete insulin when transplanted into diabetic mice. To push the research forward, Melton has co-founded a biotech company, Semma Therapeutics, and hopes to start a clinical trial in the next three or four years.

He is not alone in the race. Timothy Kieffer’s lab, in Vancouver, British Columbia, has developed another protocol to turn stem cells into beta cells that reverse diabetes in mice, and the approach is being tested by California biotech company ViaCyte. Big pharmaceutical companies are on the lookout too. AstraZeneca collaborates with Melton’s lab, and the Danish company Novo Nordisk, the worldwide leader in diabetes drugs, is working on its own project.

All have major challenges to overcome before any stem cell therapy hits the market. How results in mice translate into humans is not clear, and finding ways to trick the body’s defense system will be a huge step. New tools such as the gene-editing technology CRISPR-Cas 9 and the ability to handle huge sets of data will help, but they also raise many new questions. “The idea that we can have some mastery and control of the cells is a fantastic thing,” Melton told me, adding that while in the 20th century humans gained control over much of their physical environment, “we are now entering a century when we are going to get control over the human body.”

Chloé Hoorman, a 2016-17 Knight Science Journalism fellow, is a Paris-based journalist with Le Monde, specializing in the pharmaceutical and life sciences industries.

“You Have to Keep the Cells Happy”
Precision is critical in the lab, lest cells become damaged or experiments rendered uncertain.

Elise Engquist, a research assistant, looks at the sizes and shapes of stem cell clusters. They have to be monitored closely, as they multiply quickly and take about one month to turn into beta cells. The differentiation is done in six steps and is guided by a cocktail of growth factors added according to precise timing. Scientists always follow the same recipe, but there are still some differences between batches.
Maria Keramari, a postdoc from Greece, delicately handles a dish with human embryonic stem cells. She uses the gene-editing tool CRISPR-Cas9 to create a line of cells that will be fluorescent when they express insulin. Scientists must feed the cells at regular intervals and be careful not to hurt them during the experiment. “You have to keep the cells happy before you keep yourself happy,” Keramari says.
José Rivera-Feliciano, a postdoc from Puerto Rico, prepares samples for DNA analysis. Rivera-Feliciano works on a new family of proteins that play important roles in the development of the pancreas, and he hopes to have his own lab in the future. “It’s very competitive,” he says. “For every faculty position you have 100 to 300 candidates.”
Here, Rivera-Feliciano prepares the gel for electrophoresis, a technique for separating DNA fragments based on their size. If the result is positive, he can turn to more sophisticated and expensive DNA sequencing techniques to get more information.

“They Are Like Our Babies”
Sophisticated experiments still require some day-to-day innovation inside the lab.

Nadav Sharon, a postdoc from Israel, adds small segments of DNA, called plasmids, into petri dishes containing E. coli bacteria. Some of these bacteria will take up the foreign DNA and clone it as they replicate. Researchers routinely use these bacterial “factories” to build genetic tools that will be used to modify stem cells or create transgenic animal models. To design their experiments, scientists can dig into huge databases of bacteria, enzymes, and genes with well-known properties. Most of them can be ordered online from specialized companies and are delivered by mail.
Sharon handles vials of bacteria and puts them on ice. The heat shock causes some of them to take up the plasmids that have just been added to the vial. Sophisticated experiments often start with this kind of handmade biological tool. The recipes are well known, but there is always a measure of uncertainty in the results. Researchers can lose weeks of work without knowing what exactly went wrong.
Ornella Barrandon, a postdoc, tests 3,000 compounds to find out which ones boost the replication of beta cells in the pancreas. Beta cells derived from human embryonic stem cells could accelerate the discovery of new drugs, as research is hindered by the limited supply of beta cells extracted from cadavers. “We spend so much time on our projects, they are like our babies,” she says.

“Biology Has Taken a Turn”
But even with sophisticated genome editing tools such as CRISPR-Cas 9, this is not an exact science.

Adrian Veres, a graduate student from Canada, prepares samples for single-cell analysis, a cutting-edge technique that allows scientists to look at all the cells that constitute an organ and better understand which genes are expressed in each of them. It requires them to deal with unprecedented quantities of data: “Biology has taken a turn, and biologists must learn to deal with statistics,” Veres says. “You can’t just give the data to a statistician and look at the results.”
Chi Yang Chen, a research assistant, with a lab notebook. Each scientist has one, and must record every step of every experiment. It’s the memory and the legacy of the research done in the lab. It’s also a legal document that will be closely scrutinized in case the lab applies for a patent or is ever sued for fraud.
Jennifer Kenty, a research assistant, feeds mice with pellets. The Melton lab’s “mouse house” hosts about 1,200 mice. Beta cells derived from human stem cells are transplanted into mice that are both immunodeficient and diabetic. The goal is to determine how much insulin they secrete in response to glucose injections. Genetically modified mice are also used to test the role of different genes in the development of the pancreas and endocrine cells.
Pancreases of mouse embryos float in vials. Many questions regarding pancreas development are still open. By analyzing the type of cells that compose it at different stages of fetal development, scientists hope to create a kind of movie that will help them understand what may go wrong along the way. In other experiments, transgenic mice models are used to study the role of specific genes. Even with sophisticated genome editing tools such as CRISPR-Cas 9, this is not an exact science. “I tried once to create a CRISPR mouse,” says Nadav Sharon. “When you read the papers, it looks like magic … but that is not that simple.”