CRISPR edits lung-disease gene in utero, hitting only the affected organ in a mouse study

Companies that hope to treat severe inherited diseases via CRISPR genome editing are already testing the technique in adults, while push-the-envelope types are arguing for repairing defective genes much earlier — in IVF embryos so new they’re still in a lab dish (the “CRISPR babies” route). Now scientists in Philadelphia have taken preliminary steps toward a possible middle way: They injected CRISPR into the amniotic fluid of pregnant mice, editing a lung-disease-causing gene in a small number of mouse fetuses, they reported on Wednesday.

In utero editing offers advantages for at least some diseases, said Dr. William Peranteau, of Children’s Hospital of Philadelphia, a pediatric and fetal surgeon who co-led the study. Some mutations wreak havoc so early in development that editing genes even in a newborn would be too late: Mutations in a gene called SFTPC, whose mouse version Peranteau and his colleagues edited, cripple developing lungs so disastrously “that these kids are going to die at birth,” he said. Also, the fetal immune system is less likely than an adult’s or even a child’s to attack the CRISPR molecules or the virus that carries them.

Editing a single-cell IVF embryo, on the other hand, alters genes in every cell, which might be excessive. Such “germline” editing is also heritable, which some ethicists deem unacceptable. And embryo editing isn’t possible if conception occurs naturally, rather than in a Petri dish. Nor would it repair mutations that arise during the nine months of gestation (which can often be detected via tests of fetal DNA) rather than being present at conception.

“It’s wonderful that the field of in utero therapy is moving forward,” said Graça Almeida-Porada of Wake Forest University, an expert on fetal therapy who was not involved in the study, published in Science Translational Medicine. “For many genetic disorders, there’s not a lot that can be offered to the patient [after birth], so it’s important to develop novel therapies that provide a chance at life.” The only option for newborns with the SFTPC mutation is a lung transplant, which is rarely performed because so few tiny lungs are available for donation.

The SFTPC gene makes a protein in pulmonary surfactant. Secreted by the lungs’ alveolar cells, surfactant in humans as well as mice relaxes the surface tension in lungs so they don’t collapse with every breath. The scientists injected CRISPR into the amniotic fluid of 87 mouse fetuses on day 16 of their 20-day gestation, analogous to the third trimester in humans.

Mice can live without SFTPC, since surfactant is composed of many ingredients, said developmental biologist Edward Morrisey of the University of Pennsylvania’s Perelman School of Medicine, who co-led the study. Mutant SFTPC, however, is lethal. He and his colleagues therefore used a form of CRISPR that deletes the gene, which is relatively easy, rather than repairing it, which is harder.

Although a key challenge of CRISPR is getting it to affected cells and organs while sparing healthy ones — to cure a liver disease, it doesn’t help to CRISPR heart cells — a fetus’ lungs literally inhale it from the amniotic fluid. Few other cells picked up the genome editor.

“What’s exciting about this paper is that they showed specific targeting to just the affected organ, and even to specific cells within the lung,” said Almeida-Porada.

Every mouse born with the mutation died of lung failure within hours. But in fetuses where the gene edit succeeded (20 percent of them), seven survived for more than 24 hours. Five survived past seven days, behaving and breathing normally and hitting their growth milestones.

Mortality of 92 percent seems disappointing, to say the least. On the other hand, the scientists emphasized that “this is just a proof-of-concept study to show you can [in some cases] edit a gene in utero,” as Peranteau said. Although it will be years before such genetic surgery is ready to try in human fetuses, “we were psyched to see any survive, since normally none do.” Almeida-Porada agreed, calling even the low rate of editing and minimal survival “a great accomplishment” for a first step.

Wake Forest’s Christopher Porada, also an expert in fetal therapy (and Almeida-Porada’s husband), said the study “has implications for other genetic disorders” where it would be safer and possibly more effective to correct a gene before it makes fetal development go off the rails but only in affected organs. There is no need to repair the gene that cause Duchenne muscular dystrophy anywhere but muscles, for instance.

Last year, the Philadelphia team used another approach to edit mouse genes in utero. They injected CRISPR into pregnant mice’s vitelline vein, which drains blood from the yolk sac, and didn’t achieve organ-specific targeting as this experiment did.

Peranteau and his colleagues are working to increase the percentage of successful edits and the rate of survival, and are optimistic on both counts. The virus they chose to carry CRISPR into cells, called an adeno-associated virus, can cause lung inflammation and other problems; when it was used in normal mouse fetuses, those with healthy SFTPC, 75 percent died. A safer virus would be needed if the therapy ever moves into human studies.

In animals with longer gestation times, including people, it should be possible to get editing efficiencies much higher than the 20 percent seen in the mice, Morrisey said: The longer a fetus stays in amniotic fluid spiked with CRISPR, the greater the chance of editing its cells.

One crucial difference between people and mice would make editing SFTPC in human fetuses more difficult, however. People, unlike mice, need that gene, so CRISPR would have to repair it, not delete it, a much stiffer challenge.