Hidden Protein Interface in Potassium Channels Revealed as a Master Switch for Sleep

Sleep is one of the most universal behaviors in the animal kingdom, yet the molecular machinery that decides when a nervous system slips into its dormant state remains only partially mapped. A new study published in PLOS Genetics has now uncovered a surprisingly specific piece of that machinery: a tiny interface buried deep inside a family of voltage-gated potassium channels, where the cyclic nucleotide-binding homology domains of neighboring subunits press against one another. When that interface is disturbed, the channels open too readily, neuronal activity drops, and sleep is suppressed. The finding, reported by Xinyu Huang, Sudharsan Kannan, Gail A. Robertson, and Han Wang, elevates a subtle structural feature of channel proteins into a central player in the regulation of a complex behavior.

The team began not with a hypothesis about sleep but with an unbiased forward genetic screen in the nematode Caenorhabditis elegans, a transparent roundworm whose simple nervous system has long served as a workhorse for dissecting behavior at the level of single neurons and single molecules. Forward genetics is a deliberately blind strategy: researchers mutagenize animals, look for individuals whose phenotype has changed, and only then hunt down the responsible gene. In this case, the phenotype of interest was sleep. The worm’s sleep behavior, like that of mammals, is controlled by dedicated neural circuits, and the researchers were searching for mutations that disrupt it. Among the mutants they recovered was one whose most striking trait was a marked resistance to sleep.

Sequencing revealed that the sleep-suppressing mutation, designated G574E, sits in the gene encoding EGL-2, a KCNH-family voltage-gated potassium channel belonging to the EAG subfamily. KCNH channels are ancient and widespread: they dot the membranes of neurons and muscle cells across the animal kingdom and shape electrical excitability by letting potassium ions flow out of cells, which tends to quiet neuronal firing. What makes KCNH channels distinctive is a cytoplasmic domain at their base, the cyclic nucleotide-binding homology domain, or CNBHD, which hangs beneath the pore like a stabilizing weight. The G574E mutation alters a glycine residue within this domain, and glycine, the smallest amino acid, often serves as a flexible hinge in protein structures. Its replacement with a charged amino acid at this particular spot turned out to have dramatic consequences.

The critical insight came from structural reasoning. The glycine at position 574 does not face the channel pore or the membrane; instead, it lies at the interface where the CNBHD of one channel subunit contacts the CNBHD of the adjacent subunit. In a tetrameric channel, four such domains are arranged in a ring, and the contacts between them form a network of interfaces. The mutation disrupts this CNBHD-to-CNBHD handshake, and the result is a gain of function: the channel becomes more active than normal. For a potassium channel, more activity means more potassium efflux, more hyperpolarization of the membrane, and less excitability in the neurons that express it. In the sleep circuit of the worm, that extra dampening proved enough to keep the animal awake.

Where does this channel act? The researchers showed that EGL-2 functions cell-autonomously within ALA, a central sleep-promoting neuron in the worm. ALA is a command neuron of sorts: when sleep is called for, ALA activity helps hold the animal in its quiescent state. Using in vivo calcium imaging, which reports neuronal activity as changes in fluorescence, the team observed that the gain-of-function mutation suppresses the activity of ALA itself. This is exactly what one would expect if the mutated channel leaks more potassium: the sleep-promoting neuron falls silent, and the behavior it orchestrates, sleep, collapses with it. The result ties a molecular lesion in a single channel protein to the firing properties of a single identified neuron and from there to the behavior of the whole organism, a chain of causation that few studies in sleep biology can trace so completely.

The genetic analysis went further, testing which features of the channel are required for the phenotype. Two dependencies emerged. First, the sleep-suppressing effect of G574E requires an intact potassium selectivity filter, the exquisitely tuned structural element at the channel’s narrowest point that permits potassium ions while excluding others. This confirms that the phenotype truly reflects enhanced potassium conductance rather than some unrelated toxic effect of the mutation. Second, the phenotype requires the intrinsic ligand-occupied state of the CNBHD. KCNH channels carry a built-in ligand within their cyclic nucleotide-binding domain, an intrinsic molecule that occupies the binding pocket even though these channels are not classically regulated by cyclic nucleotides. That the CNBHD must be intact and ligand-occupied for the mutation to exert its effect underscores that the domain is not a passive appendage but an active participant in gating.

One might wonder whether the specific chemistry of the glutamate introduced at position 574 is what matters. The researchers answered this with an elegant control: they engineered a second substitution at the same glycine, G574R, which installs a positively charged arginine rather than a negatively charged glutamate. If the phenotype depended on the negative charge, G574R should behave differently. Instead, G574R also suppressed sleep. The conclusion is that it is the perturbation of the CNBHD-CNBHD interface itself, not the identity of the replacement amino acid, that unleashes the gain-of-function behavior. The glycine at this position appears to be a structural linchpin, and any sufficiently bulky or charged side chain wedged into that hinge disrupts the delicate arrangement of the domain ring.

Perhaps the most consequential finding is that this mechanism is conserved far beyond the worm. The researchers examined the residue corresponding to glycine 574 in other KCNH channels and found that mutations there produce gain-of-function effects in the ERG subfamily as well, both in C. elegans UNC-103 and, strikingly, in the human channel hERG. hERG is famous in biomedical research for its role in the heart: it repolarizes the cardiac action potential, and drugs that block it can trigger dangerous arrhythmias, making it one of the most scrutinized ion channels in pharmacology. Electrophysiological analysis of the hERG mutant G749R, carrying the analogous substitution, showed that the channel’s voltage-dependent activation shifts to more negative potentials and its activation kinetics accelerate. In plain terms, the mutant channel opens more easily and more quickly than its wild-type counterpart, a biophysical signature of a gate that has lost part of its restraint.

This conservation carries real implications for human biology and medicine. If the CNBHD-CNBHD interface restrains channel activation in hERG just as it does in EGL-2, then naturally occurring or drug-induced perturbations of this interface could alter cardiac excitability or, in neurons, neuronal excitability more broadly. The study reframes the CNBHD from a structural afterthought into a gating element whose intersubunit contacts actively hold the channel’s activation gate in check. It also adds a new entry to the short list of molecular levers that control sleep, suggesting that the excitability of sleep-promoting neurons is tuned, in part, by the mechanical integrity of a domain interface buried in the cytoplasm.

For the field of sleep research, the work demonstrates the continuing power of forward genetics to reveal unexpected mechanisms: no one set out to study CNBHD interfaces, yet the unbiased screen led directly there. For channel biophysicists, it provides a concrete structural hypothesis, that the ring of cyclic nucleotide-binding homology domains acts as a brake on KCNH activation, that can now be tested with structural and electrophysiological tools across the channel family. And for anyone interested in how brains decide to sleep, the message is that the answer can hinge on a single glycine residue, a hinge within a hinge, whose disruption silences a sleep-promoting neuron and keeps an organism awake. The worm’s restless mutants have, once again, pointed the way to a principle that extends all the way to the human heart and, quite possibly, to the human brain.

Subject of Research: Conserved CNBHD-CNBHD interface regulation of KCNH potassium channel gating and sleep

Article Title: Forward genetics reveals a conserved CNBHD-CNBHD interface that restrains EAG channel activation to regulate sleep

Article References: Huang, X., Kannan, S., Robertson, G. A., & Wang, H. (2026). Forward genetics reveals a conserved CNBHD-CNBHD interface that restrains EAG channel activation to regulate sleep. PLOS Genetics, 22(10), e1012255. https://doi.org/10.1371/journal.pgen.1012255

Image Credits: AI Generated

DOI: 10.1371/journal.pgen.1012255

Keywords: KCNH channels, EGL-2, CNBHD, forward genetics, Caenorhabditis elegans, sleep regulation, hERG, potassium channels, voltage gating, ALA neuron, calcium imaging, gain-of-function mutation