A Missing RNA-Editing Enzyme May Link Myelin Damage to Sleepless Nights
Sleep loss is often treated as a problem of brain circuits, stress hormones or disrupted daily rhythms. A new study in mice suggests that the explanation may also lie in a more structural part of the nervous system: the fatty insulation wrapped around nerve fibers. Researchers report that removing the gene Nsun5 disrupted a chemical modification on ribosomal RNA, weakened myelin in the brain’s white matter and produced profound changes in sleep. The findings offer a molecular connection between the machinery that builds proteins, the integrity of neural wiring and the ability to maintain normal sleep.
The work is especially relevant to Williams-Beuren syndrome, a rare genetic disorder caused by the deletion of a segment of chromosome 7. Most people with the syndrome also lose one copy of NSUN5, a gene encoding an enzyme that modifies ribosomal RNA. Williams-Beuren syndrome is associated with a distinctive pattern of physical and cognitive traits, as well as frequent sleep disturbances. Yet the biological pathway linking the missing gene to abnormal sleep has remained unclear. The new research, conducted by scientists at Fudan University and collaborating institutions, indicates that NSUN5 deficiency could contribute to sleep problems through damage to white matter rather than through a simple failure of the brain’s sleep clock.
NSUN5 belongs to a family of RNA-modifying enzymes that add chemical marks to RNA molecules. In this case, the enzyme installs 5-methylcytosine, or m5C, at a specific site on 28S ribosomal RNA. Ribosomes are the cell’s protein-making complexes, assembled from ribosomal RNA and proteins. They read messenger RNA and translate its nucleotide sequence into chains of amino acids. Although ribosomal RNA is sometimes regarded as a passive scaffold, chemical modifications can alter how ribosomes interact with messenger RNAs and how efficiently they translate particular groups of genes. The researchers therefore asked whether loss of Nsun5 would shut down protein synthesis broadly or instead cause a more selective imbalance in which proteins cells produce.
To investigate the question, the team created Nsun5-knockout mice and compared them with genetically normal animals. The researchers combined electroencephalography, which records electrical activity in the brain, with electromyography, which measures muscle activity, to identify sleep states and their timing. They also used RNA sequencing to measure gene transcription, ribosome profiling to determine which messenger RNAs were actively being translated, and nascent proteomics to track newly produced proteins. Conventional proteomics provided a second view of protein abundance, while transmission electron microscopy examined the fine structure of myelin. Positron emission tomography combined with computed tomography was used to map regional brain metabolism.
The first molecular result was clear: eliminating Nsun5 sharply reduced m5C modification on 28S ribosomal RNA. But the consequence was not a global collapse of translation. Instead, the knockout animals showed selective translational dysregulation. This distinction is important. A wholesale failure of ribosomes would be expected to impair nearly every cellular process and likely produce severe developmental abnormalities. Selective dysregulation implies that some messenger RNAs, or groups of related transcripts, are more sensitive than others to changes in ribosomal composition or activity. In the Nsun5-deficient mice, the affected programs were enriched for genes involved in myelination and for subunits of voltage-gated potassium channels.
Myelin is produced primarily by oligodendrocytes in the central nervous system. These specialized glial cells extend processes around axons, forming multilayered sheaths that electrically insulate nerve fibers. The insulation allows electrical impulses to travel rapidly by jumping between exposed gaps known as nodes of Ranvier, a process called saltatory conduction. If myelin becomes thin, patchy or structurally disorganized, signals can slow, fail or arrive out of sequence. The corpus callosum, the large bundle of fibers connecting the brain’s two hemispheres, was particularly affected in the knockout mice. The animals had fewer oligodendrocytes and diffuse hypomyelination, meaning that axons retained less of the normal insulating sheath.
The study also identified changes in Kcna1 and Kcna2, genes that encode voltage-gated potassium channel subunits. These channels help neurons restore their electrical state after firing and regulate the timing, duration and frequency of action potentials. Their altered translation provides a possible second route from Nsun5 loss to abnormal neural activity. A neuron communicating across a poorly insulated axon already faces problems with signal propagation; altered potassium-channel abundance could further change excitability and firing precision. Together, defects in myelin and ion-channel regulation could disturb the distributed networks that coordinate transitions between wakefulness, non-rapid-eye-movement sleep and rapid-eye-movement sleep.
The physiological effects were evident in the animals’ sleep recordings. Nsun5-knockout mice spent less time in non-rapid-eye-movement sleep and displayed altered organization of sleep during the night, when mice are normally active but also cycle through distinct sleep states. Sleep disruption was accompanied by increased regional cortical metabolic activity detected with fluorodeoxyglucose PET. Higher glucose uptake does not necessarily mean that brain function is improved; in this context, it may reflect excessive or poorly coordinated activity in cortical regions whose white-matter connections have been compromised. The result is consistent with a nervous system working harder or firing abnormally while failing to generate stable sleep.
The researchers then tested whether the phenotype could be reversed pharmacologically. Clemastine, an antihistamine that has attracted interest as a potential remyelinating drug, significantly improved sleep abnormalities in the Nsun5-knockout mice and restored the integrity of myelin. The study also tested 4-aminopyridine, a potassium-channel blocker that can improve conduction in demyelinated axons by prolonging action potentials and increasing the probability of neurotransmitter release. Treatment with 4-aminopyridine alleviated the sleep phenotype as well. Because the two compounds act through different mechanisms—one promoting myelin repair and the other modifying electrical conduction—the results support the idea that both white-matter damage and altered potassium-channel function contribute to the disorder.
The findings do not yet establish that NSUN5 loss causes sleep disruption in people with Williams-Beuren syndrome, nor do they show that either drug is a treatment for that condition. The experiments were performed in genetically engineered mice, and mouse sleep architecture, brain development and drug responses are not identical to those of humans. The study also does not determine whether myelin damage is the initiating event or whether altered neural activity and glial biology reinforce one another in a feedback loop. Nevertheless, it provides a mechanistic framework that connects an RNA modification to protein-selective translation, oligodendrocyte and myelin loss, ion-channel imbalance and abnormal sleep. The researchers’ results raise the possibility that sleep disturbance in Williams-Beuren syndrome and other white-matter disorders could be more than a symptom: it may be a measurable consequence of disrupted molecular control over the brain’s wiring and electrical communication.
