Sparse Bursts Optimize Information Transmission in a Multiplexed Neural Code

Abstract

Acknowledgments: We thank Guillaume Hennequin, Jean-Claude Béïque and Matthew Larkum for helpful discussions. We thank Loreen Hertäg, Alexandre Payeur and Stephen E. Clarke for critical reading of the manuscript as well as Greg Knoll for an independent verification of the numerical results. This work was supported by a Bernstein Award (01GQ1201) by the German Federal Ministry for Science and Education and an NSERC Discovery Grant 06872. Part of this work was conducted (RN and HS) at the Computational and Biological Learning Laboratory, Department of Engineering, University of Cambridge, UK.Author Declarations: The authors declare no conflict of interest.Keywords: neural coding |cerebral cortex | bursting | multiplexing | decoding | spike timing | short-term plasticity Author Contributions: HS and RN designed the study and wrote the manuscript. RN performed the simulations analyzed the results and carried out the theoretical analysis. peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/143636 doi: bioRxiv preprint first posted online Jun. 30, 2017; AbstractMany cortical neurons combine the information ascending and descending the cortical hierarchy. In the classical view, this information is combined nonlinearly to give rise to a single firing rate output, which collapses all input streams into one. We propose that neurons can simultaneously represent multiple input streams by using a novel code that distinguishes single spikes and bursts at the level of a neural ensemble. Using computational simulations constrained by experimental data, we show that cortical neurons are well suited to generate such multiplexing. Interestingly, this neural code maximizes information for short and sparse bursts, a regime consistent with in vivo recordings. It also suggests specific connectivity patterns that allows to demultiplex this information. These connectivity patterns can be used by the nervous system to maintain optimal multiplexing. Contrary to firing rate coding, our findings indicate that a single neural ensemble can communicate multiple independent signals to different targets.

Learning from invariants predicts upcoming behavioral choice from spiking activity in monkey V1

“…It appears likely that bursty neurons described here correspond to the recently reported class of bursting excitatory neurons in the V1 of the macaque, suggested to have an important role in encoding and transmission of cortical signals [37]. Alternatively, a top-down choice signal could drive bursty neurons in the superficial layer through dendritic feedback inhibition, as suggested by a theoretical study [47].…”

Distinct Properties of Layer 3 Pyramidal Neurons from Prefrontal and Parietal Areas of the Monkey Neocortex

“…Thus, DLPFC B-L3PNs may specialize in transmitting top-down signals to other cortical regions. Burst firing might also enhance information transmission by spike trains in neuronal ensembles, enabling PN groups to simultaneously process top-down and bottom-up streams of information (Naud and Sprekeler, 2018). Thus, the higher proportion of burst-firing L3PNs observed here in DLPFC might equip DLPFC neuronal ensembles with a richer repertoire of information-processing mechanisms.…”

Burst-dependent synaptic plasticity can coordinate learning in hierarchical circuits

“…Integration of credit-carrying feedback signals should be close to linear and avoid saturation (i.e., feedback signals should be linear with respect to any credit information). Experimental and theoretical work have addressed steering [12,55], multiplexing [56][57][58][59], alignment [34,41,60,61] or linearity [62] in isolation. , often by learning in an offline fashion [34-37, 40, 41, 63, 64], without learning rules based on spikes [28,30,[35][36][37]65] or without learning to solve tasks that necessitate hierarchical processing.…”