The research, led by Maria Florencia Acutain, Fernanda Mariana Silva, Luisina Castellari, and María Verónica Baez at the Instituto de Biología Celular y Neurociencia in Buenos Aires, focused on N-methyl-D-aspartate receptors (NMDARs), the glutamate-gated ion channels that sit at the heart of learning and memory. These receptors are tetramers built from two obligatory GluN1 subunits and two regulatory subunits whose identity changes dramatically across development. In immature circuits, the GluN2B subunit predominates; as networks mature, GluN2A becomes enriched, altering channel kinetics, temporal integration windows, and the thresholds that govern synaptic plasticity. This developmental switch has made GluN2A a prime suspect in the search for the molecular determinants of adult cognition.

Yet the evidence linking GluN2A to spatial behavior has long been clouded by a methodological problem. Most studies have relied on constitutive knockout models, in which the gene encoding GluN2A is silenced from the earliest stages of embryonic development. While these animals show impairments in spatial tasks, contextual fear conditioning, and object location memory, it has remained impossible to determine whether the deficits arise from the absence of GluN2A in the adult brain or from compensatory rewiring that occurs during development. To sidestep this confound, the Argentine team turned to a viral strategy: they injected an adeno-associated vector carrying a short hairpin RNA (shRNA) designed to knock down GluN2A expression specifically in the CA1 region of the dorsal hippocampus of adult Wistar rats, well after the developmental GluN2B-to-GluN2A switch had completed.

The surgical procedure was precise. Under ketamine and xylazine anesthesia, rats received bilateral injections of either the GluN2A-targeting vector or a scrambled control vector, delivered at a multiplicity of infection of 2 × 10^4 in two microliters per side. Fourteen days later, once the knockdown had taken hold, the animals entered a battery of behavioral tests designed to probe different facets of hippocampal function: the novel object location task, an adapted three-chamber social test, a two-trial social memory test, and both a direct and a reverse Barnes maze. All sessions were recorded and analyzed by researchers blinded to treatment, and after testing, the CA1 region of each hippocampus was microdissected to confirm that GluN2A mRNA levels had indeed been reduced.

The first surprise came from the novel object location task, a paradigm that relies on a single training session and depends entirely on hippocampal function. When one of two identical objects was displaced to a new position, control rats of both sexes spent significantly more time exploring the moved object than the stationary one — the classic signature of spatial memory. GluN2A-knockdown rats, however, showed no such preference. Neither the males nor the females distinguished between the displaced and the familiar object, and their discrimination indices failed to reach the learning criterion that both control groups passed with ease. The deficit was not a matter of reduced exploration: total object exploration time was comparable across all groups, and the animals showed no bias toward the left or right object during training.

But when the researchers gave the animals more time — four consecutive training sessions on the Barnes maze, a low-stress spatial task that is neither emotionally nor appetitively driven — a striking sex difference emerged. GluN2A-knockdown males, despite failing the single-session object location task, learned the Barnes maze across repeated sessions, progressively shortening their latency to find the escape hole just as control animals did. GluN2A-knockdown females, by contrast, showed persistently elevated latencies through the first three sessions and only began to improve on the fourth. In the test session, control animals of both sexes and knockdown males spent significantly more time in the target quadrant than in the opposite one, but knockdown females distributed their exploration evenly across all four quadrants, indicating weak learning of the escape hole location.

The reversal phase of the Barnes maze — in which the escape hole was moved 180 degrees to a new position — exposed a second layer of impairment that affected both sexes. All groups learned the new location during training, showing similar latencies regardless of sex or treatment. Yet in the test session, only control animals showed a clear preference for the new target quadrant. Both male and female GluN2A-knockdown rats spent similar amounts of time in all quadrants, failing to demonstrate the cognitive flexibility needed to update their spatially guided behavior when the rules of the environment changed. This finding was unexpected, because cognitive flexibility is typically attributed to cortical structures rather than to hippocampal NMDAR signaling, and it suggests that the hippocampus and prefrontal cortex may work in concert to support this function.

To dissect the exploration strategies underlying these deficits, the team analyzed the behavioral patterns that rats displayed on the Barnes maze, classifying each movement as direct, serial, or random search. Using Markov chain analysis of the transitions between these patterns, they found that control animals maintained stable transition probabilities across both the direct and reverse phases of the task. GluN2A-knockdown animals did not. Knockdown females were the only group that showed random-to-random transitions and never performed direct-to-direct sequences, exploring the maze in a fundamentally more disorganized manner. Knockdown males, meanwhile, changed their strategy profile between the two phases of the task, shifting how they moved from one search pattern to the next. These altered exploration dynamics — not a simple loss of motivation or locomotor capacity — appear to underlie the impaired performance in spatially guided tasks.

Crucially, the deficits were not global. In social tasks, GluN2A-knockdown rats of both sexes socialized normally, recognized novel conspecifics in the structured three-chamber paradigm, and interacted with unfamiliar animals with latencies and total exploration times comparable to controls. Only in a free-exploration social context — the two-trial social memory test, where animals move freely rather than being confined — did knockdown rats fail to discriminate between a familiar and a novel partner. This dissociation is telling: it suggests that GluN2A is not required for social recognition per se, but rather for the rapid spatial updating that free exploration demands. Consistent with this interpretation, the researchers found that GluN2A expression levels correlated with discrimination indices only in tasks with demanding spatial components, and not in tasks with fixed spatial references.

The implications reach beyond basic neuroscience. Reduced GluN2A expression has been linked to schizophrenia-related symptoms, and impaired cognitive flexibility is a recognized feature of the disorder — one that typically emerges in adolescence or early adulthood, after the developmental GluN2B-to-GluN2A switch has been established. The authors propose that subtle reductions in GluN2A expression in mature hippocampal neurons may selectively compromise the precision and flexibility of spatial behavior, contributing to the cognitive inflexibility seen in neuropsychiatric and neurodegenerative conditions. They also caution that their study has limitations: the knockdown approach cannot distinguish between deficits in acquisition, consolidation, and retrieval, the viral targeting was verified by qPCR rather than histological validation, and group sizes were relatively small. Still, by isolating GluN2A function in the adult brain — free from developmental confounds — the work establishes this receptor subunit as a key modulator of spatially guided behavior and highlights the importance of considering sex as a biological variable in studies of hippocampal cognition.

Subject of Research: The role of hippocampal GluN2A NMDA receptor subunits in spatially guided behavior and cognitive flexibility in adult rats

Article Title: Hippocampal GluN2A knockdown differentially alters spatial exploration and behavioral performance in male and female rats

Article References: Acutain, M. F., Silva, F. M., Castellari, L., & Baez, M. V. (2026). Hippocampal GluN2A knockdown differentially alters spatial exploration and behavioral performance in male and female rats. Frontiers in Behavioral Neuroscience, 20, Article 1935028. https://doi.org/10.3389/fnbeh.2026.1935028

Image Credits: AI Generated

DOI: 10.3389/fnbeh.2026.1935028

Keywords: GluN2A, NMDA receptors, hippocampus, spatial learning, cognitive flexibility, Barnes maze, sex differences, AAV-shRNA knockdown, CA1, synaptic plasticity, schizophrenia, behavioral neuroscience