Fig. 5: Schematic of dendritic non-linear computation in PV+ interneurons during SPW-Rs and SPW-R doublets.
From: Sharp-wave ripple doublets induce complex dendritic spikes in parvalbumin interneurons in vivo
Our model comprises four supralinear layers, which covert information from dendritic inputs (xi) to the somatic output (output AP) during the two events of SPW-R-doublets (SPW-R 1st, SPW-R 2nd). Layer I: local dendritic activation above (NMDA-AMPA spikes) and below (SPW-R-EPSPs) the threshold of NMDA-AMPA spikes. Layer II: dendritic Ca2+ spikes (SPW-R-dSpikes), which appear on top of NMDA-AMPA spikes. Layer III: SPW-R doublet-associated dSpikes (doublet-dSpike). The inset in layer III shows a truth table for computation. Layer III works as a temporal coincidence detector: it is activated and generates large responses (doublet-dSpike) when both inputs (WI and WII) are active (1, 1) with a time delay (Δt) during SPW-R doublets. With all other input combinations [(1, 0), (0, 1), (0, 0)], it simply transmits the dendritic responses of layer I and layer II: the SPW-R-EPSPs and SPW-R-dSpikes.
