Interactions between the cortex and the thalamus are essential for major brain functions such as sensory information processing and integration, sleep and wake regulation and cognitive processes. The thalamic reticular nucleus (TRN) is strategically positioned within the thalamocortical circuit and has a strong inhibitory control over the thalamus. It can act on a global scale, such as suppressing the flow of sensory information from the thalamus to the cortex during sleep. The TRN also acts locally on the activity of single cells or small cell groups. To reconcile both of these global and local aspects of TRN functions, we studied the cellular, synaptic and functional heterogeneity of the TRN, with a focus on the comparison between the classical sensory TRN and the less well-described limbic TRN.
In study 1, using anatomical tracing and cellular electrophysiology, we identified the dorsal presubiculum (dPreS), the retrosplenial cortex (RSC) and the anterior thalamic nuclei (ATN) as part of a novel thalamo-cortical circuit involving the limbic TRN in mice. The dPreS, RSC and ATN are three key structures for spatial navigation. dPreS/RSC excitatory glutamatergic synapses formed on TRN and ATN are part of a feedforward circuit through which TRN-mediated inhibition generates large burst-mediated inhibitory synaptic currents. The PreS/RSC afferents to the TRN showed driver-like characteristics, which is unprecedented for corticoreticular synapses and expands the scope of the TRN heterogeneity to the nature of its synaptic afferents. We further investigated the role of the limbic TRN in the control of head-direction neurons that were previously described to be located in the anterodorsal thalamus. The width of the tuning curve of head-direction neurons in the thalamus was broadened upon chemogenetic silencing of the TRN, revealing a novel form of internal sensory gating by the TRN. About half of the head-direction neurons showed action potential discharge patterns consistent with feedforward inhibitory responses upon light activation of dPreS/RSC. These data suggest that the limbic TRN sharpens the tuning of thalamic head-direction neurons under dPreS/RSC control. Finally, we investigated the potential function of the limbic TRN in the hidden version of the Morris watermaze. We discovered that chemogenetic silencing of the limbic TRN biased the search patterns towards allocentric strategies and generated perseverance to previously learned escape positions, suggesting an impairment of the egocentric system in which the head-direction system plays a critical role.
In study 2, we combined opto-tagging of TRN sectors with in vitro electrophysiological recordings and discovered that the limbic TRN neurons produced less repetitive burst firing than their sensory counterpart. The burst discharge of sensory TRN neurons is known to generate sleep spindles that propagate to the cortex, that are a marker of sleep quality and that correlate with memory consolidation. Consistently, local field potential recordings in the prefrontal cortex that is related to the less bursty limbic TRN revealed smaller amplitude and slower sleep spindles compared to sensory ones, making the heterogeneity of the TRN a critical player in local sleep rhythms.
This thesis summarizes elements supporting the heterogeneity of the TRN, in particular between the sensory and the limbic TRN. It also provides a novel function for the limbic TRN in the spatial navigation system.