Neural Networks

Understanding the neonatal brain 

For the past two decades, Frances Jensen, MD, of the Department of Neurology and Program of Neurobiology at Children’s, has been learning why most infants seem immune to anticonvulsants and developing new treatments specific to their biology.

Beginning in the laboratory, she studied the brain’s molecules and modeled epilepsy in animals. Then, with Children’s Pathology Department, Jensen made corresponding observations in human tissue. Collectively, her findings revealed just how different the physiology and biochemistry of a baby’s brain are from an adult’s. “It’s practically a different species,” Jensen says.

Adult brains are balanced between two states, excitation and inhibition. But Jensen and others have shown that the rapidly developing infant brain is more inclined toward excitation.

“A baby’s increased brain activity is designed to create new connections that are the underpinnings of learning,” Jensen explains. “But this turns out to be a double-edged sword. There are certain diseases, such as epilepsy, that are caused by over-activation of the brain.”

Most anticonvulsant drugs—developed for adults—increase inhibition. Yet the developing brain doesn’t have many of the inhibitory synapses that the medications target. “This physiologic difference explains why current medications are not effective for infants,” Jensen says.

Jensen returned to the lab with this knowledge. In animal models of early-life seizures, she found that blocking a certain kind of excitatory receptor—something rarely tried, since most drugs target inhibitory receptors—prevented further seizures and stopped all of their long-term consequences.

The drug she used, called topiramate, has been approved by the FDA to control seizures in adults and in children over age 3. Unfortunately, it doesn’t yet exist in IV form and will take several years to develop for clinical trials.

Seeking other options, Jensen worked with pediatric neurologist Kevin Staley, MD, chief of Pediatric Neurology at Massachusetts General Hospital, to explore the other half of the equation: how inhibition might be boosted in infants to block seizures.

Conventional anticonvulsants mimic the action of GABA, a natural inhibitory messenger, by activating GABA receptors on the surface of brain cells. In adult cells, this opens channels so chloride moves into the cell, giving it a negative charge that makes it less excitable and inhibits seizure activity. But in babies’ cells, chloride concentration is already high, so when GABA receptors open, chloride flows out of the cell, toward the area of low concentration. This wrong-way chloride flow creates a paradoxical excitatory reaction that may actually worsen seizures.

“An infant’s GABA receptors are actually doing the opposite of what they do in adults,” Jensen says. “Their brains wants so much to be excitable, they’re even using classical inhibitory receptors to bring about excitation.”

But how? Jensen and Staley found the answer in two molecules that regulate cells’ chloride levels. One, called KCC2, transports chloride out of cells; the other, NKCC1, brings chloride in. In adult rats, KCC2 predominates in nerve cells, keeping internal chloride concentrations low; thus, when GABA receptors are activated, chloride comes in, with an inhibitory effect. But in newborn rats, they found very little KCC2.

Examining brain tissue from babies and young children who had died, they found the same pattern in humans: KCC2 was initially absent in the upper part of the brain where the seizures originated, but rose over the first year of life. Conversely, NKCC1 levels were high during the fetal and newborn periods, falling during the first year of life.

“NKCC1 is expressed unopposed in the immature brain, likely keeping cellular chloride levels high,” says Jensen. “We thought that blocking NKCC1 and its inward transfer of chloride, might get immature neurons to act like older neurons and give GABA a chance to do its job.”

Fortuitously, research had shown that an existing drug called bumetanide blocks NKCC1 in the kidney. If it did the same in the brain, perhaps it could keep chloride levels low inside newborns’ nerve cells and allow them to respond to anticonvulsants. A trial in baby rats confirmed the idea, showing that bumetanide successfully blocked seizures. Even better, when combined with the anticonvulsant phenobarbital, which works poorly when given alone, bumetanide was even more effective.