09/09/2025 - Press release
A study published in Science identifies the complementary role of the main families of neurons in representing the environment.
A multidisciplinary team coordinated by Dr. Manuel Valero, head of the Neural Computation Laboratory at the Hospital del Mar Research Institute, in collaboration with New York University (NYU) and with the participation of the Instituto de Neurociencias de Alicante (INA) and Cardenal Herrera-CEU University, has developed an artificial intelligence tool capable of classifying neurons according to their electrical activity based on their response to light. The study, published in the journal Science, has revealed how different families of interneurons cooperate in building the brain's internal orientation system, which we use to represent the environment and learn how to navigate through it.
The researchers analyzed the behavior of more than 7,000 neurons from the hippocampus and the cerebral cortex of mice while they performed a navigation task. With this tool, they were able to assign each neuron a cellular identity according to its family of origin. The analysis shows that these neuronal families perform specific and complementary functions: some regulate the precision of the map, others its stability, and others enable it to adapt to changes in the environment.
"This line of work is redefining our understanding of the brain: not as a tabula rasa that simply records what comes through the senses, but as a system that produces actions based on circuits shaped by evolution and refined by learning," explains Dr. Manuel Valero, coordinator of the project and head of the Neural Computation Laboratory at the Hospital del Mar Research Institute.
The study is the result of collaboration with the groups of Dr. Bernardo Rudy and Dr. György Buzsáki at New York University, international leaders in the study of neuronal diversity and brain oscillations. The team used genetically modified animal models in which each family of neurons expressed a light-sensitive protein, which made it possible to selectively activate these neuronal groups and record their activity during spontaneous behavior and rest. These data were then used to train an artificial intelligence system capable of distinguishing between different neuronal families based on brain activity recordings.
"We have developed a tool to characterize neuronal diversity from electrophysiological recordings. By incorporating this diversity into our models of the hippocampal circuit, we improve our understanding of how this region enables the formation of orientation maps during learning," notes Dr. Pablo Abad Pérez, postdoctoral researcher and first author of the study.
The tool developed is available in open access and can now be applied to the study of other brain regions. The next step for the research team is to use it to analyze how neuronal circuits are altered in diseases such as Alzheimer's, epilepsy, major depression, or Down syndrome, with the aim of advancing towards more specific therapeutic strategies.
Reference article
Valero, M.; Abad-Pérez, P.; Gallardo, A.; Picco, M.; García-Hernández, R.; Brotons, J.; Martínez-Félix, A.; Machold, R.; Rudy, B.; Buzsáki, G.; et al. Cooperative actions of interneuron families support the hippocampal spatial code. Science 2025, 389(6764), eadv5638. https://doi.org/10.1126/science.adv5638
© Institut Hospital del Mar
d'Investigacions MèdiquesLegal Notice and Privacy Policy | Cookie Policy | Site Index | Accessibility | Find Us | Contact
