Spinal muscular atrophy treated in utero in mice

Researchers have taken preliminary steps toward treating a devastating neurodevelopmental disorder, spinal muscular atrophy, with a genetic therapy delivered in the womb, before the worst damage to motor neurons can occur. 

The scientists at the University of California, San Francisco, and Johns Hopkins University in Baltimore injected an antisense oligonucleotide, or ASO, into the amniotic fluid of pregnant mice, and from there it traveled into the brains and spinal cords of mouse fetuses with the genetic disorder and substantially improved the motor function of the resulting pups. 

The research, detailed Wednesday in Science Translational Medicine, is a proof of concept. More work would need to be done before such an approach could be tested in human fetuses. But the research teams are working with the California biotech Ionis — whose ASO-based medicine Spinraza was approved for use in newborns with SMA in 2016 — with a goal of doing so in the not too distant future. 

Before 2016, being born with the most severe form of SMA came with a two-year death sentence. The disease is caused by mutations in the SMN1 gene, which produces a protein essential for motor neuron survival — without it they die. That loss can be compensated for by a related gene, called SMN2, which produces smaller amounts of the protein. The more copies of SMN2 a person has, the milder the symptoms are likely to be. But about 60% of SMA patients have two copies or fewer, resulting in rapid progressive loss of muscle control until even the small muscles between the ribs and inside the nose and mouth that push air into and out of the lungs fail — usually before an affected individual’s second birthday.  

Spinraza and other therapies that have been approved in the last decade, including the gene therapy Zolgensma and an oral pill called Evrysdi, have transformed the lives of SMA patients and their families. Today, newborns across the U.S. are genetically screened for the disease, and if they have it, immediately started on medication; the earlier treatment can start, the better the outcome. For SMA patients with the most severe form of the disease, however, these treatments, while improving their condition, are not cures. 

The problem, scientists have found, is that there is an explosive wave of motor neuron die-off that happens months before any appearance of disease. In fact, this mass cell death event often starts in the womb. “There’s clear evidence this disease begins in utero,” said Charlotte Sumner, a professor of neurology and neuroscience at Johns Hopkins who co-led the new research. Her group had previously found that levels of a cytoskeletal filament that neurons release into the blood when they die were massively elevated right after birth and before any symptoms appeared. 

“The neurons are just dying like crazy,” Sumner said. “So there’s this rapid wave of neurodegeneration early on that we have to get ahead of because once it happens there’s no turning it back.”

Years ago, Sumner began exploring various ways to get ahead of the wave. That work accelerated when she met Tippi MacKenzie, a fetal surgeon at UCSF who is pioneering the emerging field of in utero genetic medicine. Their groups teamed up and began evaluating whether it would be safe and effective to administer existing treatments even earlier in an SMA patient’s life. 

Working with sheep, which offer a better model of human development than mice, MacKenzie’s team first looked at a gene therapy like Zolgensma, which uses a viral vector to deliver its therapeutic piece of DNA. Using a reporter gene to see where it would wind up, they observed that the genetic instructions went where they had hoped — to the brain and spinal cord and muscle cells — but also, in lesser amounts, to locations where they shouldn’t, including the ovaries of the female lambs. 

That presented a potential problem; regulators have been strident that genetic therapies demonstrate an infinitesimal risk of altering the human germline — cells that pass down genetic instructions to future generations. So the groups moved on to the ASO approach, which involves small pieces of genetic material that bind to RNA molecules and either block or modify their protein-making activity. They began working with Ionis, the company that had developed an ASO drug for SMA that was later licensed by Biogen and marketed as Spinraza. 

Unlike gene therapy, which adds a new, healthy copy of the SMN1 gene, the ASO developed by Ionis is a short string of genetic material that binds to the product of the SMN2 gene, in a section known to recruit molecules that direct it to generate more of the functional SMN protein. In infants and children, it’s given as an intrathecal injection directly into the liquid surrounding the spinal cord so that it can reach the motor neurons. 

For their latest study, the groups looked at two delivery methods of the ASO. Working with a mouse model of a severe form of SMA, the Hopkins team injected the treatment into the amniotic fluid of pregnant mice. They observed that it migrated into the central nervous systems of the developing pups, where it elevated levels of the healthy SMN protein and led to better motor function and increased survival of the animals, compared to pups that were given the treatment after birth. 

Because sheep don’t have an SMN2 gene, the UCSF team worked with healthy fetal lambs to establish if the ASO could safely get into the animals’ central nervous systems at therapeutic concentrations. The researchers injected some of the animals with the treatment directly into a space at the back of their skulls where cerebrospinal fluid flows from the brain to the spinal cord (mimicking intrathecal injections in humans) and others directly into the amniotic fluid of their mothers’ wombs. 

The scientists found that therapeutic concentrations of the drug were reached in the fetal lambs’ central nervous system, or CNS, more consistently with the direct brain injections than with the ASO-spiked amniotic fluid.

“The data is really exciting to see that you get a reasonable biodistribution to the CNS and that it was safe and well-tolerated,” Sumner said. However, she and her co-authors acknowledged that more work needs to be done to improve the odds of successful delivery via the amniotic fluid route, which would be a much less invasive procedure.