A steady flow of ions across cellular channels is crucial for heart and brain health.

Much like a city, the body relies on electricity to power its functions. Electrical currents keep the heart beating on pace, relax and contract muscles, and flow from neuron to neuron. To modulate the flow of electricity, the body uses voltage-gated ion channels, which open briefly to allow sodium ions to pass through and then close immediately. If the gates close improperly, ions can leak through and disrupt healthy rhythms.

For example, mutations that result in overexpression of the genes SCN5A and SCN8A, which code for two sodium channels, can cause ion leakage that leads to cardiac arrhythmia and epilepsy. Existing drugs for these conditions tamp the flow of ions through the gate, but sometimes they can reduce the flow too much, leading to other issues.

In a study published recently in Cell, researchers instead devised a new molecule to nudge the gate closed, opening the door to better therapeutics for these conditions.¹

Biophysicist Manu Ben-Johny at Columbia University, who is the senior author on the work, came up with the idea to generate a novel peptide after seeing a social media post from Sergey Ovchinnikov, a coauthor on the paper. Ovchinnikov had described a way to design synthetic peptides with Google Colab, an online platform for running code and accessing the high-speed computers needed for deep learning. Using the platform, Ben-Johny and his colleagues generated a peptide that would bind to the existing intracellular mechanism for closing the gate. They dubbed the resulting chain of 21 amino acids: engineered late-current inhibitor X by inactivation-gate release (ELIXIR). Unlike drugs that bind to the sodium channel to block ion flow, ELIXIR instead targets the gate-closing mechanism.

Heart muscle cells with mutations in SCN5A leak far more ions than control cells, Ben-Johny and his team found. Modifying the cells to express ELIXIR greatly reduced the electricity levels, though they were still elevated over the controls.

In neurons and skeletal muscle cells with SCN5A and SCN8A mutations, ELIXIR shortened action potential duration to near the level of wildtype cells. The peptide had no effect on control cells, suggesting it’s not active when the sodium channels are functioning typically.

“It’s a beneficial thing to be able to have some specificity for pathogenic channels,” Ben-Johny said. “It’s a feature that we didn’t expect to get from the get-go.”

That feature bodes well for treating people with SCN8A-related epilepsy, who generally have one healthy copy and one mutated copy of the gene, said neurogeneticist Miriam Meisler at the University of Michigan, who was not involved in the new work. Meisler discovered the SCN8A channel in 1995 and characterized the first human mutation in the gene in a patient with epileptic encephalopathy in 2012.^2,3

SCN8A is a “Goldilocks gene,” she said: The body needs just the right amount to function properly. If the peptide also dampened the healthy copy, levels would drop too low, creating new problems.

“The very unusual fact that this compound seems to only act on the mutant gene would make it a very desirable drug,” Meisler said. Ben-Johny’s team should test how long the effects of ELIXIR last, she said.

ELIXIR also normalized electrical activity in cardiac muscle cells from a mouse with a mutation in SCN5A, as well as in heart cells grown from stem cells of a person with arrhythmogenic long-QT syndrome 3, a disorder resulting from mutations in SCN5A.

Next, the researchers plan to test whether ELIXIR works on other mutant sodium channel variants and reduces seizures and arrhythmia in the mouse models. They hope to develop a suite of molecules that can target specific ion channel dysfunctions, Ben-Johny said.

“Our work illustrates how computational protein design can be so powerful,” he added. “With improved accessibility, it is possible for anyone to use these methods to develop proteins that have never existed in nature with functions that are beneficial to humanity.”