The research team, led by Katsunori Kobayashi of Wakayama Medical University and Nippon Medical School together with Hiroki Imbe, set out to answer a deceptively simple question: can the maturation status of a developing neuron be reversed by excessive activity, the way it can in adults? The stakes are high. Immaturity of the granule cells of the dentate gyrus — the hippocampal gateway for new information — has been documented in a remarkable range of mouse models of neuropsychiatric disease, from schizophrenia models lacking Schnurri-2 or alpha-CaMKII to epilepsy models with SNAP-25 mutations. If hyperexcitation can derail maturation during development, it could provide a common pathogenic mechanism linking seizures, inflammation and aberrant receptor signaling to later psychiatric illness.

To probe this, the researchers used electroconvulsive treatment, or ECT, as a controlled model of neuronal hyperexcitation. Importantly, they were not modeling electroconvulsive therapy for depression; they were using brief bilateral electrical shocks — 15 milliamperes for one second at 100 pulses per second — delivered through ear-clip electrodes to trigger generalized seizures. Each mouse received three rounds of ECT over four or five days, with the identical stimulus parameters applied at every age from two weeks to nine weeks. The animals were anesthetized with isoflurane to prevent sudden unexpected death in epilepsy, and none died under the treatment conditions.

The team’s yardstick of neuronal maturity was frequency facilitation at the mossy fiber synapses, the output connections that dentate granule cells send to the CA3 region. Frequency facilitation is a presynaptic form of short-term plasticity: when synapses are stimulated repeatedly at low frequency, transmitter release progressively strengthens. Mature granule cells show marked frequency facilitation, and this signature rises sharply during postnatal development, reaching near-adult levels by about four weeks. In adult mice, both chronic antidepressant drugs and repeated ECT reduce this facilitation back to the juvenile level — the hallmark of what the group has named “dematuration,” a pseudo-immaturity of otherwise mature neurons.

When the same ECT paradigm was applied to developing mice, the results were strikingly age-dependent. At three weeks of age, ECT had no significant effect on mossy fiber synaptic transmission. But by five weeks, frequency facilitation was significantly reduced, and the effect first became detectable at four weeks, reaching the magnitude seen in adult ECT-treated mice by six weeks. Crucially, the researchers ruled out trivial explanations: the baseline synaptic efficacy, measured by the ratio of field EPSPs to presynaptic fiber volleys, did not change with ECT, and paired-pulse facilitation, another short-term plasticity measure, was also unaffected. The reduction in frequency facilitation therefore could not be explained by a simple increase in baseline transmitter release probability, which generally suppresses facilitation. Instead, it pointed to a genuine reversal of the granule cells’ maturation status.

Molecular marker analysis confirmed the electrophysiological picture. In six-week-old mice, ECT strongly reduced the expression of Tdo2, the gene encoding tryptophan 2,3-dioxygenase, a well-characterized marker of mature granule cells. The expression of Calb1, encoding the calcium-binding protein calbindin, was not changed under the present stimulus conditions — likely because the current intensity was lower than in previous adult studies and because calbindin downregulation requires more than ten ECT rounds to persist, so the expression most likely recovered by the time of sampling 24 hours after the last treatment. Importantly, ECT did not alter the expression of Dcx, a marker of neuronal progenitors and early immature neurons, or Map2, a general neuronal marker, excluding the possibility of non-specific gene suppression. The inducibility of dematuration, the authors conclude, emerges by postnatal week four.

What changes in the brain to open this window? The researchers suspected N-methyl-D-aspartate, or NMDA, glutamate receptors, which are required for ECT-induced dematuration in adults. Recording from the medial perforant path, the major excitatory input onto granule cells, they found that both AMPA receptor-mediated and NMDA receptor-mediated synaptic responses, normalized to the presynaptic fiber volley, increased significantly during postnatal development. In other words, the glutamatergic input to the granule cells becomes more efficacious as the animals mature — and this maturation of synaptic efficacy paralleled the developmental emergence of ECT responsiveness.

The team then asked how granule cells at different ages respond to high-frequency synaptic activation, the kind of input that ECT and seizures are thought to impose. High-frequency stimulation of the perforant path induced only transient post-tetanic potentiation in adult slices, with no lasting change. In young mice, the post-tetanic potentiation was larger, and — remarkably — the same stimulation induced a small long-term depression, especially in mice aged two weeks or younger. This LTD was blocked by the NMDA receptor antagonist D-APV, indicating that NMDA receptor-dependent signaling activated by high-frequency input changes fundamentally during the first weeks of postnatal life. The authors speculate that in young animals, NMDA receptor activation levels may simply be insufficient for dematuration, or that the intracellular signaling pathways linking NMDA receptors to the dematuration machinery — candidates include calcineurin, alpha-CaMKII, Cdk5 and glycogen synthase kinase-3, all of which have been implicated in granule cell maturation — are not yet fully developed.

The developmental timing carries translational weight. Comparing rodent and human brain development, the period from postnatal week two to week nine in mice roughly corresponds to infancy through adolescence in humans. Calbindin expression in the mouse dentate gyrus reaches mature levels around postnatal week five, whereas an adult-like calbindin pattern in the human dentate gyrus is observed at about 11 years of age. If similar activity-dependent plasticity operates in the human dentate gyrus, the authors suggest, neuronal hyperexcitation might induce dematuration of human granule cells around the age of 10 years or older. Intriguingly, the onset of idiopathic generalized epilepsies peaks around age 15, and status epilepticus during postnatal development is known to increase the lifetime incidence of epilepsy. Because dematured granule cells exhibit increased somatic excitability, the induction of dematuration could theoretically promote subsequent epileptic discharge and the transition to chronic epilepsy — a self-reinforcing loop that new therapies might break.

The findings also speak to the NMDA receptor hypothesis of schizophrenia, which originated from the observation that ketamine, an NMDA receptor antagonist, induces schizophrenia-like symptoms in humans. The present work implies that the story may be bidirectional: NMDA receptor hypofunction during development could impair proper maturation, while NMDA receptor hyperfunction might promote aberrant dematuration. In support of the latter, mutant mice with dentate gyrus-specific NMDA receptor hyperfunction show working memory deficits, a cognitive symptom of schizophrenia. Neuroinflammation, which can increase neuronal excitability and has been documented in Schnurri-2 knockout mice, may further facilitate dematuration during the vulnerable window. The authors note limitations: only male mice were studied, and the ECT current was not normalized to body size, so age-dependent differences in seizure induction may have influenced the results. Still, the central conclusion stands — the plasticity that links neuronal hyperexcitation to a reversal of maturation is itself a product of development, established around postnatal week four in mice, and the signaling cascades that drive it are potential therapeutic targets for preventing the pathogenesis of neuropsychiatric disorders.

Subject of Research: A mouse study showing that activity-dependent neuronal dematuration in the dentate gyrus becomes inducible from postnatal week four, coinciding with maturation of NMDA receptor signaling.

Article Title: Developmental changes in activity-dependent plasticity of maturation-related neuronal phenotypes

Article References: Kobayashi, K., & Imbe, H. (2026). Developmental changes in activity-dependent plasticity of maturation-related neuronal phenotypes. Frontiers in Synaptic Neuroscience, 18, Article 1926357. https://doi.org/10.3389/fnsyn.2026.1926357

Image Credits: AI Generated

DOI: 10.3389/fnsyn.2026.1926357

Keywords: neuroscience, dentate gyrus, hippocampus, neuronal maturation, dematuration, electroconvulsive treatment, NMDA receptors, synaptic plasticity, mossy fiber synapses, epilepsy, schizophrenia, neuropsychiatric disorders