Enabling precise cell-type-specific dissection of orientation and memory circuits in retrosplenial cortex

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT In humans, damage to a brain region called the retrosplenial cortex leads to pronounced spatial disorientation and severe retrograde and anterograde memory deficits. Similar navigational and memory impairments are also seen in rodents with either lesions or chemogenetic inactivation of the retrosplenial cortex. Despite its critically important functions, the cells, circuits, and computations of the retrosplenial cortex remain understudied, especially when compared to those of the hippocampus and entorhinal cortex. We have recently shown that a small, excitable, pyramidal neuron – only found in layers 2/3 (L2/3) of the granular retrosplenial cortex (RSG) – has properties that are very different from its more standard (regular-spiking; RS) neighbors and is uniquely suited to computing compass-like directional information over long durations. We have named this neuron the Low Rheobase (LR) cell. Using optogenetic ex vivo circuit mapping, we have subsequently found that inputs from the thalamus (source of directional information) and from the dorsal subiculum (source of spatial information) converge selectively onto these small LR cells, while avoiding nearby RS cells. Thus, LR neurons are ideally positioned to support the RSG’s spatial orientation computations. During non-REM sleep, hippocampal ripples (known to be important for memory consolidation) are propagated, via the dorsal subiculum, to L2/3 of the RSG. Since these dorsal subicular outputs selectively recruit LR but not neighboring RS cells, LR neurons are also ideally positioned to play a central role in the consolidation of memories from the hippocampus to the RSG. Despite this strong rationale to dissect the behavioral role of LR cells in vivo, two critical hurdles remain to enable a TargetedBCP R01 submission in the near future. First, it is technically challenging to electrophysiologically record from large numbers of simultaneous LR cells in vivo. This is because they are located within a narrow ~120 um band of RSG tucked away close to the midline, with vertical access prevented by blood vessels. Second, the transcriptomic signature of LR neurons remains unknown, preventing the rational selection or production of transgenic mouse lines that selectively and specifically label LR neurons. To overcome these hurdles, in Aim 1, we will develop and test custom-designed probes optimized to record large numbers of L2/3 RSG neurons. In Aim 2, we will utilize Allen Brain Institute databases, 10x Next-Gen sequencing, and Patch-seq to identify the transcriptomic class corresponding to the morphophysiological class of LR neurons. The completion of these Aims will set the stage for a subsequent TargetedBCP R01 submission that will utilize large-scale recordings and causal opto/chemogenetics to carefully decipher the importance of LR neurons in the representation and consolidation of spatial information.