Learning and attention increase visual response selectivity through distinct mechanisms

Jasper Poort,Katharina A Wilmes,Antonin Blot,Angus Chadwick,Maneesh Sahani,Claudia Clopath,Thomas D Mrsic-Flogel,Sonja B Hofer,Adil G Khan

doi:10.1016/j.neuron.2021.11.016

Abstract

SummarySelectivity of cortical neurons for sensory stimuli can increase across days as animals learn their behavioral relevance and across seconds when animals switch attention. While both phenomena occur in the same circuit, it is unknown whether they rely on similar mechanisms. We imaged primary visual cortex as mice learned a visual discrimination task and subsequently performed an attention switching task. Selectivity changes due to learning and attention were uncorrelated in individual neurons. Selectivity increases after learning mainly arose from selective suppression of responses to one of the stimuli but from selective enhancement and suppression during attention. Learning and attention differentially affected interactions between excitatory and PV, SOM, and VIP inhibitory cells. Circuit modeling revealed that cell class-specific top-down inputs best explained attentional modulation, while reorganization of local functional connectivity accounted for learning-related changes. Thus, distinct mechanisms underlie increased discriminability of relevant sensory stimuli across longer and shorter timescales.

Highlights

Head-fixed mice ran through a virtual approach corridor (Figure 1A) where the walls displayed a short stretch of circle patterns followed by gray walls for a random distance chosen from an exponential distribution (Figure 1C, top)
We found that during learning, SOM cells become de-correlated from pyramidal, PV, and vasoactive intestinal peptide (VIP) neurons, with the largest changes between cell classes
Modeling response changes during learning and attention What changes in network properties underlie the observed changes during learning and attention? We recently developed a multivariate autoregressive (MVAR) linear dynamical system ll Article model to predict the activity of single cells based on interaction weights with their local neighbors

Summary

Introduction

Learning and attention both selectively enhance the processing of behaviorally relevant stimuli (Gdalyahu et al, 2012; Goltstein et al, 2013; Li et al, 2008; McAdams and Maunsell, 1999; Ni et al, 2018; Reynolds and Chelazzi, 2004; Rutkowski and Weinberger, 2005; Schoups et al, 2001; Speed et al, 2020; Wiest et al, 2010; Yan et al, 2014; Yang and Maunsell, 2004). When animals learn what sensory features are task relevant or when they focus their attention on task-relevant features, early sensory cortical representations often undergo substantial changes. Activity modulations during learning and attention are not uniformly distributed throughout the neural population but are restricted to subsets of neurons (see, for example, Chen et al, 2008; McAdams and Maunsell, 1999; Poort et al, 2015; Schoups et al, 2001; Yan et al, 2014). Both learning and attention lead to sharper and more distinct information being sent to downstream regions though subnetworks of learning- or attentionmodulated cells

Results

Discussion

Conclusion