Elaine Fuchs was disappointed with the slow pace of conventional mouse genetics. To understand how skin stem cells develop and renew, she and her colleagues had been engineering strains of mice in which they could selectively turn off one or more genes. But the process, which requires breeding several generations of mice, was taking years.
Despite more than two decades of effort, Fuchs, an HHMI investigator at Rockefeller University, and her team managed to create only 30 engineered mouse strains—a small fraction of the genes they’d like to study. When a stem cell becomes active, for example, changes occur in the expression of roughly 200 genes. “There are still a myriad of genes up- or down-regulated whose role in stem cell biology is unknown,” says Fuchs. Making mutant mice to study all of them could take a lifetime.
Fuchs wanted a faster way. And last year, after several frustrating failed attempts, she found one. The technique bypasses breeding altogether. Using a virus and a tiny needle, she and her colleagues can now do in a few weeks what used to take a year or more. “It has opened up doors for us,” she says. “We can analyze genetic pathways [in mice] in ways previously possible only in flies and worms.”
The discovery has been years in the making. About a decade ago, scientists figured out how to alter lentiviruses—round, spiky viruses like HIV that cause chronic infection—to carry small, hairpin-shaped strands of RNA into mammalian cells. Scientists can engineer these RNA strands to block expression of particular genes, a technique called RNA interference. Fuchs thought the viruses might be a quick way to block gene expression in skin stem cells.
The technique worked well in skin cells growing in culture. But in an animal, skin tissue is made up of many layers. “When applied to the skin’s surface, the lentivirus went in, but only into the very top layer of dead cells,” Fuchs says. No matter what she and her colleagues tried—roughing up the skin, injecting the virus with a tattoo needle—they couldn’t get the lentivirus to go where they wanted: into the stem cells.
Then, around 2008, the team—Fuchs, postdoc Slobodan Beronja, postdoc Scott Williams, and graduate student Geulah Livshits—had a breakthrough idea. Why not try to infect mouse embryos rather than adult mice? Early in development, the embryos are covered by skin that is just a single cell layer thick. As development proceeds, these early skin stem cells generate the stratified epidermis and its hair follicles. If the researchers could inject the lentivirus into the amniotic sac, they speculated that the virus might infect this early cell layer and then pass on its genetic cargo to all those cells’ progeny. If so, the team would be able to selectively block expression of any gene in the skin in just a few days.
First, the team anesthetized a mother mouse. Then they carefully made an incision to access the embryos. The team relied on ultrasound to guide a glass needle—thinner than a human hair—into the pea-sized amniotic sacs. Once the needle was in place, they injected the virus. The virus dispersed throughout the amniotic fluid, infecting only the cells it touched—those covering the embryo’s surface.
The team had to time it just right, injecting the virus when the embryos were nine-and-a-half days old. Before then, the virus may infect cells that move inward to form the central nervous system. Wait too long, however, and the embryo develops a temporary protective covering, so the virus can’t reach the skin stem cells.
To test their technique, the researchers loaded the lentivirus with a gene that contained the blueprints for a fluorescent protein. After the pups were born, the team put them under a fluorescence microscope. Some of the pups’ skin glowed. The infection had worked. “That was really, really exciting,” Livshits says. And the effect was lasting—the mice still expressed the fluorescent protein as adults. The researchers also found that when they injected more virus, a larger proportion of skin cells became infected.
The researchers still didn’t know whether they could use the virus to block gene expression in the mice. So they engineered the lentivirus to hold an RNA hairpin meant to block the gene for alpha catenin, a protein that helps cells stick together, and injected it into mouse embryo amniotic sacs.
“It worked,” Livshits says. The protein was largely undetectable in cells that had been infected, while still present at normal levels in the uninfected cells. And the infected tissue looked identical to tissue taken from a mouse unable to express alpha catenin in the skin. The team reported their results in the July 2010 Nature Medicine.
The technique may work for other tissues as well. Christer Betsholtz, a biologist at the Karolinska Institute in Stockholm, Sweden, hopes to tweak the technique to study vascular development. Rather than injecting the virus into the amniotic sac, he plans to inject it into the embryo’s heart or a large blood vessel. The hope is that the virus will travel through the circulatory system and infect the cells that line the blood vessels, Betsholtz’s research interest. But the heart is a much smaller target than the embryonic sac. Delaying the injection a few days might make the injection more feasible. “This will not be easy, but it should be doable,” he says.
Since their discovery, Fuchs and colleagues have used their new technique to investigate the molecular mechanisms behind asymmetric division, a phenomenon in which a cell divides into two different kinds of progeny cells rather than two identical daughters.
Fuchs has always been intrigued by the fact that when the skin begins to stratify, it changes its angle of cell division from parallel to perpendicular relative to the embryo’s body. She and her colleagues speculated that this spindle reorientation might play a key role in asymmetric division. Evidence from fly studies shows that a specific set of genes regulates asymmetric cell division, enabling the mother cell to partition proteins to the two daughters unequally. Only one of the daughter cells receives proteins involved in a signaling pathway called “Notch.” This pathway helps determine whether the daughters become tissue or stem cells. “We wondered whether this ancient mechanism might also be used to regulate skin progenitors in the mouse,” Fuchs says.
To test this hypothesis, she and her team infected embryos with lentiviruses that reduced expression of three key proteins involved in this process. In the absence of these proteins, the perpendicular divisions were perturbed, and the skin didn’t stratify as it should. To test whether the skin lacking these proteins formed a tight seal, the researchers dropped the embryos into blue dye. Embryos with normal skin exclude the dye, but these embryos adsorbed it, turning blue. The team reported their results in the February 17, 2011, issue of Nature.
“The amount of genetics that it would have taken to do this set of experiments with conventional techniques would have been...not insurmountable, but it would have taken a very long time—not something desirable for any postdoc aiming to start their own laboratory,” Fuchs says.
The technique has “really taken off” in the Fuchs lab, Livshits says. She estimates that 10 lab members are using it to study everything from sweat glands to corneal epithelium—and, of course, the skin and its cancers. Rather than focusing on proteins one by one as they used to do, researchers can now focus on many proteins in a genetic pathway, gaining more insights in less time. “For a skin biologist like myself, it’s heaven,” Fuchs says. “Now we can put the molecular pieces into our biological puzzles at a pace not thought possible even a few years ago.”