Title : Astrocytic GPCR signaling in the hippocampus modulates neuronal computation underlying inferential fear memory through L-lactate in mice
Abstract:
Humans have the ability to generalize information learned from one experience and apply it to another situation. Both the hippocampus and anterior cingulate cortex (ACC) play key roles in the encoding of new information using established associative memory schema, but it remains unclear whether this brain region computes cognitive short-cuts to support associative inferential learning and memory. We expanded the Pavlovian fear conditioning paradigm to test our hypothesis that following paired auditory-visual stimuli training, direct visual-foot shock fear conditioning can generate indirect auditory inference fear memory.
In this study, mice performed experiments using two indirectly related elements (sound and foot shock) linked through an overlapping element (light) via direct sound-light and light-shock associations, providing the foundation for transitive inference. By recording neuronal calcium activity with jGCaMP7s, we show that recruitment of the hippocampus during reasoning increases with the relational demands required for the reasoning process. Moreover, during the inference test, auditory stimulation evoked marked visual cortex activation, indicating that reactivation in the visual cortex supports auditory inferential fear memory retrieval. Computational analysis of in vivo electrophysiological recordings revealed that dCA1 neurons, including those projecting to the ACC, exhibited activity patterns consistent with the retrieval of learned associations, while ACC neurons generated a prospective representation of the inferred threat. These findings were supported by enhanced theta-band coherence between dCA1 and ACC during successful inference. Optogenetic silencing of dCA1 or V1 impaired inference memory retrieval, confirming the necessity of these regions.
We tested our hypothesis that astrocytic activation in dCA1 is necessary for developing inferential auditory fear. Chemogenetic activation of astrocytic Gq signaling in dCA1 enhanced inferential fear memory and was associated with increased L-lactate levels, whereas activation of astrocytic Gi signaling impaired inference memory and reduced extracellular L-lactate. Specific knockdown of astrocytic EAAT2 impaired inference memory, and this deficit was rescued by local administration of exogenous L-lactate. Conversely, the memory-enhancing effect of astrocytic Gq activation was blocked by pharmacological inhibition of EAAT2.
In conclusion, our data suggest that memory encompasses not only direct experience but also information integrated from multiple independent events. ACC-projecting hippocampal neurons contribute to inferential reasoning by integrating new experiences into existing memory networks, with dCA1 and ACC performing dissociable computational roles during inference. Astrocytic G-protein-coupled receptor signaling, glutamate transporter EAAT2, and L-lactate play key roles in modulating the neural computations underlying inferential fear memory.
