Soma location of parvalbumin-expressing interneurons defines their contribution to feedforward and feedback inhibition
Authors/Creators
- Ocaña-Fernandez, Maria del Angel (Data collector)1
- Loureiro, Rita (Data collector)2
- Hertäg, Loreen (Data curator)3
- Tamm, Johanna (Data collector)1
- Vida, Imre (Project member)2
- Sauer, Jonas-Frederic (Project member)4
- Elgueta, Claudio (Data collector)1
- Sprekeler, Henning (Project member)3
- Bartos, Marlene (Contact person)1
- 1. Institute for Physiology I, Medical Faculty, University of Freiburg, 79104 Freiburg, Germany
- 2. Institute of Integrative Neuroanatomy, Charité, Universitätsmedizin Berlin
- 3. Modelling of Cognitive Processes, Technical University of Berlin, 10587 Berlin, Germany
- 4. Center for Integrative Physiology and Molecular Medicine (CIPMM), Saarland University, 66421 Homburg, Germany
Description
Feedforward and feedback inhibition are two elementary interrelated processes that are integrated in cortical microcircuits. Using in vitro whole-cell recordings of parvalbumin-expressing interneurons (PVIs) in the mouse dentate gyrus (DG), we provide first evidence that the soma location of PVIs defines their preference for providing feedforward or feedback inhibition to the network. PVIs with somata at the outer border of the granule cell to molecular layer (PVIouters) are more strongly recruited by afferent entorhinal cortex inputs, providing short-latency uniform feedforward inhibition to granule cells as compared to PVIs with soma location at the inner granule cell layer to hilar border (PVIinners). In contrast, PVIinners are more strongly recruited by granule cells than PVIouters due to their stronger synaptic plasticity and higher input connectivity, preferentially providing a delayed, reliable feedback inhibition to the DG network. In vivo single-unit recordings of optogenetically identified PVIs revealed that activity of both subtypes differentially relate to dentate spikes, with PVIouters being predominantly active during the first half and PVIinners preferentially discharging during the second half of dentate spikes. Our data-driven computational model provides an explanation on how the functional integration of both PVI subtypes in the DG network support pattern separation processes by promoting winner-takes-all mechanisms.
Notes
Files
Codes_NatureCommunications.zip
Files
(5.2 GB)
| Name | Size | Download all |
|---|---|---|
|
md5:9de9bf6838dd03aebd1de7a9bc0b1643
|
5.2 GB | Preview Download |