ETD

• escape behaviour
• perineuronal nets
• retrosplenial cortex

Perineuronal nets (PNNs) are specialized reticular structures of the extracellular matrix (ECM) that ensheath neurons, particularly GABAergic parvalbumin-positive (PV) interneurons, across the entire mouse and human brain. The formation of PNNs occurs as one of the last acts of neural development and by aggregating around the somata of neurons they progressively contribute to the structural and functional stabilization of neuronal circuitries. Besides developmental plasticity, PNNs also contribute to synaptic stabilization, neuroprotection, ion homeostasis, and the preservation of long-term memories.
In recent work, our research group has performed a systematic brain-wide analysis of their distribution in the mouse brain, which highlighted that PNNs are particularly abundant in the retrosplenial cortex (RSP). However, the role of PNNs in this area is still not well understood. The RSP is critical for spatial navigation as it integrates the representation of spatial landmarks with the egocentric subject coordinates. Recent evidence showed that in the presence of a threatening stimulus, neurons in the RSP and the superior colliculus develop a sharp tuning for the direction of a previously identified safe shelter which is necessary to trigger a fast and precise escape. Given the known role of PNNs in circuit plasticity, we speculate that they might be relevant for establishing, maintaining and updating this shelter-direction encoding.
Therefore, in the present thesis we aimed at investigating this hypothesis by experimentally manipulating PNNs in the context of RSP-dependent escape behavior.
We built a setup consisting of a 90 cm arena equipped with a shelter that can be moved around the perimeter, a camera, two speakers, and a PC running a MATLAB script through which we provided the aversive auditory stimulus. To assess baseline responses, mice were first exposed to few presentations of the aversive auditory stimulus, after a short period of habituation to the arena. To test whether PNNs were involved in the stability of shelter direction encoding, mice underwent intracortical injection of either chondroitinase ABC, an enzyme capable of degrading chondroitin sulfate proteoglycans, or a vehicle solution in the RSP. The removal was then verified through PNN staining. Finally, to investigate whether PNNs are involved in the plasticity of the shelter representation circuit, mice were exposed to the same test after the closure of the original shelter and the opening of a new one positioned at a 90-degree angle.
We quantified escape behavior using two complementary metrics: a shelter orientation index measuring directional preference at escape initiation, and a shelter distance index assessing final positioning relative to the correct (open) shelter. Preliminary analysis suggests that removal of PNNs is not sufficient to disrupt mice accuracy in the orientation phase of escape behavior, but might instead make the animal more adaptable in the context of changing environmental stimuli.
In conclusion, while these findings would require further experimental validation, this work suggests that PNNs may not be strictly required for maintaining circuit stability upon environmental changes, potentially indicating a role in modulating the reliability of behavioral responses.