Cell-specific wiring routes information flow through hippocampal CA3

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

The hippocampus is dogmatically described as a trisynaptic circuit. Dentate gyrus granule cells, CA3 pyramidal neurons (PNs), and CA1 PNs are serially connected, forming a circuit that critically enables memory storage in the brain. However, fundamental aspects of hippocampal function go beyond this simplistic ‘trisynaptic’ definition. CA3 PNs connect not only to CA1, but also establish the largest autoassociative network in the brain. In addition, CA3 PNs are not uniform, differing in their morphology, intrinsic properties, and even the extent of granule cell input. Understanding how these different subtypes of CA3 PNs are embedded in the hippocampal network is essential for our quest to understand learning and memory. Here, we performed simultaneous multi-cellular patch-clamp recordings from up to eight CA3 PNs in acute mouse hippocampal slices, testing 3114 possible connections between identified cells. Combined with post-hoc morphological analysis, this allowed full characterization of neuronal heterogeneity in functioning microcircuits. We demonstrate that CA3 PNs can be divided into distinct ‘deep’ and ‘superficial’ subclasses, with altered input-output balance. While both subtypes formed recurrent connectivity within classes, connectivity between subtypes was surprisingly asymmetric. Recurrent connectivity was abundant from superficial to deep, but almost absent from deep to superficial PNs, thereby splitting CA3 into parallel recurrent networks which will allow more complex information processing. Finally, we observed innervation of PN subclasses by distinct interneurons, a potential mechanism to gate information flow through CA3 sublayers. Together, our data present a major revision to the classical ‘trisynaptic’ view of the hippocampus, bringing us closer to understanding its complex action in information storage.

Article activity feed