Researchers at IN-CSIC/UMH in Spain corrected a single gene’s expression in one population of amygdala neurons in adult mice and reversed their anxiety-like behavior, according to a 2025 study in iScience. The finding, led by first author Alvaro Garcia under the direction of Juan Lerma, challenges the assumption that anxiety rooted in abnormal brain development is permanently wired in. Instead, it points to an ongoing synaptic imbalance that can be fixed well after the circuit has already formed.
Why the BLA-to-CeL circuit matters for anxiety research right now
The amygdala is not a single structure. It contains distinct subregions that pass signals in a specific sequence. The basolateral amygdala (BLA) sends excitatory projections into the central amygdala, where a subpopulation of neurons in the centrolateral division (CeL), called regular firing neurons, helps gate fear and anxiety responses. When that relay weakens or strengthens abnormally, emotional behavior shifts. Prior optogenetic work showed that stimulating BLA terminals in the central amygdala could produce acute, reversible reductions in anxiety in mice, proving the pathway has a direct causal role.
The new iScience paper builds on that foundation with a different question: if a genetic defect disrupts BLA-to-CeL communication from birth, can repairing it in adulthood still work? The researchers used mice carrying extra copies of the Grik4 gene, which encodes the GluK4 kainate receptor subunit. Excess GluK4 had already been shown to alter synaptic transmission and produce autism-related behavioral traits in earlier work from the same research line, including a study in which Grik4 overexpression changed excitatory–inhibitory balance and led to social and cognitive deficits. Separately, complete loss of GluK4 through genetic knockout also changed anxiety and depression measures in mice, as reported in Molecular Psychiatry. Together, those studies established that GluK4 levels must stay within a narrow range for normal affective behavior. Too much or too little, and the circuit misfires.
That dose sensitivity is what makes the new result so specific. The team did not broadly dampen anxiety circuits or flood the brain with a drug. They normalized Grik4 expression in one cell type, BLA pyramidal neurons, using a stereotaxic injection of AAV-Cre-GFP virus. The intervention left the rest of the brain untouched. The fact that this single correction restored synaptic strength onto CeL regular firing neurons and reversed anxiety-like behavior supports a pointed interpretation: the anxiety phenotype in these mice is maintained by an active synaptic imbalance, not by a structural defect locked in during early development.
Gene dosage, viral rescue, and what the iScience data show
The study, titled “Central role of regular firing neurons of centrolateral amygdala in affective behaviors,” used a conditional genetic strategy. Grik4-overexpressing mice carried a version of the gene flanked by loxP sites, allowing Cre recombinase to cut and normalize the extra copies. By injecting AAV-Cre-GFP directly into the BLA of adult animals, the researchers restricted the gene correction to pyramidal cells in that region. They then measured two things: whether synaptic drive onto CeL regular firing neurons recovered, and whether the animals’ behavior on standard anxiety tests changed.
Both measures shifted. Functional communication between BLA pyramidal cells and CeL regular firing neurons was restored after the intervention. Electrophysiological recordings showed that excitatory postsynaptic currents in CeL regular firing neurons, which had been weakened by Grik4 overexpression, returned toward control levels once the excess gene dosage was cut back in the BLA. This synaptic rescue was not observed in neighboring cell types that did not receive the viral correction, underscoring that the effect was anatomically and genetically specific.
Anxiety-like behavior, assessed through established paradigms such as time spent in the open arms of an elevated plus maze and exploration of the center of an open field, decreased to levels comparable with control animals. Grik4-overexpressing mice without the viral rescue avoided exposed areas and showed heightened anxiety-like responses. After the BLA-targeted gene normalization, those same mice behaved more like wild-type controls, suggesting that the repaired BLA-to-CeL communication was sufficient to rebalance affective output.
The study identified CeL regular firing neurons as the specific downstream target whose input had to be repaired for the behavioral rescue to occur. When the researchers examined other CeL neuron subtypes, the synaptic changes were less tightly linked to anxiety behavior. That level of circuit specificity is unusual. Most anxiety research manipulates broad receptor classes or entire brain regions. Here, the correction targeted one gene in one cell population and tracked the effect through a defined synapse onto a defined neuron type.
The research line behind these findings stretches back through multiple publications. A Cell Reports study from the same group had previously linked Grik4 overexpression to circuit-level output imbalances and disease-relevant behaviors, showing that excessive GluK4 in limbic pathways disrupts the balance between excitation and inhibition and alters social interaction, cognitive flexibility, and stress responsiveness. The 2025 iScience paper narrows the mechanism further by isolating the BLA-to-CeL leg of that disrupted circuit and showing it is both necessary and sufficient for the anxiety phenotype, at least in this mouse model.
Open questions after the BLA pyramidal cell rescue
The strongest implication of the study is also its biggest unresolved question. If adult-only gene normalization can reverse anxiety in mice whose circuits developed abnormally, how far does that principle extend? The Grik4-overexpression model recapitulates features of autism spectrum disorders and anxiety, but human anxiety disorders involve many genes, environmental factors, and brain regions beyond the amygdala. Whether correcting a single synaptic relay in one circuit would have measurable effects in a more genetically complex organism is unknown.
Another open issue is timing. The mice in this study received the viral rescue in adulthood, but after a lifetime of altered amygdala signaling. The fact that behavior could still be shifted argues that, at least for this pathway, maladaptive anxiety is not entirely cemented by early developmental wiring. However, the work does not yet define a critical period or determine whether even later interventions, or briefer ones, would be equally effective. Longitudinal experiments with varied ages and dosing schedules will be needed to map how plastic the BLA-to-CeL circuit remains across the lifespan.
Specificity also cuts both ways. Correcting Grik4 dosage in BLA pyramidal cells produced a clean rescue in this model, but it is not clear whether similar precision can be achieved in the human brain. Stereotaxic viral injections into the amygdala carry surgical risks, and current gene therapy vectors cannot yet guarantee restriction to one projection-defined neuron class. Off-target effects in neighboring cells or connected regions could, in principle, introduce new imbalances even as they fix old ones. Translating such a targeted approach into a therapy would require far more selective delivery systems and rigorous safety data.
Finally, the study focuses on anxiety-like behavior but touches on broader questions about comorbid traits. Grik4 overexpression models aspects of autism spectrum disorders, and earlier work from the same group linked the gene to social and cognitive alterations as well as anxiety. The new data show that normalizing BLA output to CeL regular firing neurons can correct anxiety measures, but they do not yet clarify which other behavioral domains, if any, are similarly reversible via this pathway. Future experiments could test whether parallel corrections in other circuits, or combined interventions across multiple projection targets, are needed to address the full spectrum of traits associated with GluK4 dysregulation.
Even with these caveats, the study delivers a clear conceptual message: some forms of pathological anxiety may be maintained by ongoing, adjustable synaptic states rather than by fixed developmental lesions. By pinpointing a single gene in a defined set of amygdala neurons and showing that its adult correction can restore both synaptic strength and behavior, the work from IN-CSIC/UMH offers a concrete example of how circuit-level precision can reshape affective outcomes. For basic neuroscience, it underscores the value of mapping not just which regions are involved in anxiety, but which specific neuron types and connections sustain it over time. For translational research, it hints that future interventions, whether genetic, pharmacological, or neuromodulatory, might achieve lasting relief by rebalancing key synapses in already-formed circuits rather than attempting to rebuild them from scratch.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
