A popular blood pressure drug also appears to reduce the stress of being born and curb symptoms of autism in rodents. Though critics say the findings may not apply to people, the work could reveal new understandings—and treatments—for the complex neurodevelopmental disorder.
Our brains are naturally hyperactive in utero, at least according to rodent studies. High levels of chloride inside neurons keep these cells buzzing with electrical activity, which may promote rapid brain development and growth, says Yehezkel Ben-Ari, a neuroscientist at the biomedical research agency INSERM’s Mediterranean Institute of Neurobiology in Marseille, France, and acting CEO of Neurochlore, a company that develops treatments for developmental disorders. During labor, however, the release of a hormone called oxytocin causes chloride ion levels in cells to drop, calming these neurons down. That, in turn, appears to help babies deal with the stress of childbirth, Ben-Ari suggests.
Because complications during labor have been associated with higher rates of autism, Ben-Ari and his colleagues speculate that this protective switch may not occur for babies who go on to develop the disorder. The team tested its hypothesis in two separate rodent models of autism. One group of mice had the most common genetic mutation associated with the human disorder—it’s seen in 2% to 6% of people with autism—while a group of rats was exposed in utero to sodium valproate, an epilepsy drug known to significantly increase risk of autism in children whose mothers take the medication. Although there’s no way to create a truly “autistic” animal, these mice and rats have social deficits similar to those seen in people with the condition: Some lack interest in new companions, for example, and don’t react—squeak in distress—as much when separated from their mothers.
Using electrodes to measure the flow of current in and out of neurons of the rodent fetuses in utero and after they were born, the researchers found that chloride levels were higher than average in the embryonic brain cells of the "autistic" animals, and stayed that way through adulthood. EEG tests showed abnormally large and powerful oscillations in the rodents' hippocampi, where memories are processed and stored, suggesting that brain activity was not being properly controlled. Normal rats and mice, in contrast, had high levels of chloride inside their neurons in utero, but these dropped soon after delivery and their brain activity appeared normal.
Next, the team tested whether bumetanide, a drug that blocks chloride transport channels in neurons and is frequently prescribed for high blood pressure, could lower chloride and restore normal neuron function in the autistic rodents. In 2012, the team had conducted a small clinical trial with the drug, finding that it modestly improved social behaviors in children with mild autism . Ben-Ari and his colleagues laced the drinking water of pregnant mice and rats with the bumetanide, then once again measured their offspring's neuronal chloride levels and excitability before and after birth. Chloride levels and brain activity were restored to approximately normal levels in both groups , the researchers report online today in Science. In addition, adult autistic mice treated with the drug resumed more normal social behaviors.
Ben-Ari believes the new work underscores the value of ongoing clinical trials of bumetanide in children who have already begun to show symptoms of autism. He does not foresee that the drug will be administered prenatally to pregnant mothers, however, as there is no way to diagnose autism in the womb.
Others remain cautious: Because human brains develop at vastly different rates from rodent brains—3 to 4 weeks of development in people is equivalent to a day or less in mice, for example—the team's observations may not hold in human autism, notes Emanuel DiCicco-Bloom, a neuroscientist at Rutgers University’s Robert Wood Johnson Medical School in Piscataway, New Jersey.
Larry Young, a neuroscientist at Emory University in Atlanta, shares DiCicco-Bloom's concern that the authors of the new study are too quick to equate rodent models with humans, but says it is "very exciting" to see evidence that hormonal signaling between mother and child during birth can make changes to brain circuitry that last a long time.
Overall, the study is "pretty awesome," DiCicco-Bloom says. “I don’t think that anyone would have predicted" that chloride levels would be abnormally high in both models of autism, or that just one treatment with bumetanide before birth could correct them, he says. Although there is no way to diagnose or treat autism before birth in humans, he says, a host of obstetrical complications have been associated with higher risk of autism and intellectual disability. "This could be one of the mechanisms by which that occurs," he says."
*Correction, 6 February, 3:55 p.m.: This item has been modified to reflect the fact that both mice and rats were used in the autism models.