Stimulus-Selective Response Plasticity in Primary Visual Cortex: Progress and Puzzles

Abstract

Stimulus-selective response plasticity (SRP) is a robust and lasting modification of primary visual cortex (V1) that occurs in response to exposure to novel visual stimuli. It is readily observed as a pronounced increase in the magnitude of visual evoked potentials (VEPs) recorded in response to phase-reversing grating stimuli in neocortical layer 4. The expression of SRP at the individual neuron level is equally robust, but the qualities vary depending on the neuronal type and how activity is measured. This form of plasticity is highly selective for stimulus features such as stimulus orientation, spatial frequency, and contrast. Several key insights into the significance and underlying mechanisms of SRP have recently been made. First, it occurs concomitantly and shares core mechanisms with behavioral habituation, indicating that SRP reflects the formation of long-term familiarity that can support recognition of innocuous stimuli. Second, SRP does not manifest within a recording session but only emerges after an off-line period of several hours that includes sleep. Third, SRP requires not only canonical molecular mechanisms of Hebbian synaptic plasticity within V1, but also the opposing engagement of two key subclasses of cortical inhibitory neuron: the parvalbumin- and somatostatin-expressing GABAergic interneurons. Fourth, pronounced shifts in the power of cortical oscillations from high frequency (gamma) to low frequency (alpha/beta) oscillations provide respective readouts of the engagement of these inhibitory neuronal subtypes following familiarization. In this article we will discuss the implications of these findings and the outstanding questions that remain to gain a deeper understanding of this striking form of experience-dependent plasticity.

Keywords: cortical plasticity; experience dependent plasticity; habituation; novelty detection; primary visual cortex; visual recognition memory.

FIGURE 1

Manifestations of stimulus-selective response plasticity (SRP). (A) In a standard SRP experiment, mice are chronically implanted with tungsten microelectrodes in L4 of the binocular region of V1. They are then head-fixed and passively view black-and-white phase-reversing grating stimuli in order to elicit visually-evoked potentials (VEPs). (B) In juvenile (P25) mice, a single day of stimulus presentation leads to substantial potentiation of the VEP, which is maintained across subsequent days of viewing. The VEP elicited by a novel (N) stimulus orientation on the test day is reduced relative to the VEP elicited by the familiar (F) orientation. Adult (P60) mice require several days to reach asymptotic SRP, but after 6 days of stimulus presentation display a F/N difference similar to that of juvenile mice. Replotted from Schecter et al. (2017). (C) SRP also manifests as an increase in the peak firing rate of multi-unit activity recorded from L4 of V1. As with VEPs, the peak firing rate is potentiated for the familiar orientation relative to the novel orientation. Replotted from Cooke et al. (2015). (D) SRP is also apparent in changes in the spectral power of various oscillation frequencies. Relative to familiar stimuli, novel stimuli elicit greater high-frequency (gamma) oscillations and diminished low-frequency (alpha/beta) oscillations. This can be observed as a difference in the normalized power within these frequency ranges both when comparing the first and last days of SRP induction, and when comparing responses to familiar and novel orientations presented within the same session. Replotted from Hayden et al. (2021). (E) SRP manifests as a decrease in the activity of L4 principal cells when measured with calcium imaging during blocks of F stimuli. Mice expressing GCaMP6f in L4 principal cells are implanted with a prism, and a 1-p microscope is used to measure cellular fluorescence during SRP induction. Average dF/F responses decrease over the course of SRP induction, and are greater when viewing novel stimuli relative to familiar stimuli. White asterisk in this panel indicates location of implanted prism. ΔF/F calculated as F_stim – F_gray/F_gray. Replotted from Kim et al. (2020). Black asterisks throughout this figure indicate significant differences.

FIGURE 2

Stimulus-selective response plasticity and orientation-selective habituation (OSH) are input-specific and require NMDARs in V1. (A) A piezoelectric device situated beneath the forepaws of a mouse during a standard SRP experiment can be used to measure movements of the animal elicited by stimulus onset. The raw voltage recording is rectified and then normalized to the pre-stimulus onset period. The area under the curve is averaged to determine the magnitude of the visually evoked fidget (the “vidget”). (B) Over the course of a typical SRP experiment, the magnitude of the vidget diminishes as the animal becomes familiar with the same stimulus across days. Presentation of a novel stimulus elicits a larger vidget relative to presentation of the familiar stimulus, demonstrating the stimulus-selectivity of habituation. (C) Changes in the vidget follow a similar timecourse to SRP, but in the opposite direction of change observed for VEPs. (D–F) Both SRP and OSH are eye specific. (D) Distinct orientations were presented monocularly to each eye. On test day, VEPs and vidgets were measured in response to monocular presentation of the stimulus orientation familiar to the eye (blue), the orientation familiar to the opposite eye but novel to the viewing eye (yellow), and an orientation novel to both eyes (red). (E) Relative to the familiar stimulus for each eye, the vidget evoked by both novel stimuli were significantly increased, and the VEPs evoked by both novel stimuli were significantly decreased. There were no significant differences between the responses to the stimuli that were novel to either eye and the true novel stimuli. (G–I) Both SRP and OSH require NMDARs in V1. (G) Grin1^fl/fl mice were injected with a virus expressing either Cre recombinase with a GFP tag (KO) or a control virus expressing GFP alone (Ctrl). (H) Ctrl mice show normal OSH, but KO mice do not display a significant familiar-novel difference for vidgets. (I) Ctrl mice show normal SRP induction (days 1–6) and expression (day 7 test), but both are disrupted in KO mice. Figure replotted from Cooke et al. (2015). Significant comparisons are denoted throughout by an asterisk and non-significant comparisons by n.s.

FIGURE 3

Inhibition plays a role in SRP expression. (A) Mice expressing Cre recombinase in either PV+ or SOM+ inhibitory neurons were injected with a Cre-dependent GCaMP7f virus in V1. Activity in PV+ or SOM+ cells in L4 was then recorded using a 2-photon microscope. (B) Over the course of SRP induction, PV+ cells showed a significant decrease in stimulus-evoked fluorescence, eventually displaying suppression below baseline levels. PV+ activity for N stimuli was also significantly increased relative to activity for F stimuli. (C) SOM+ cells showed the opposite pattern of changes to PV+ cells, with an increase in stimulus-evoked fluorescence over the course of SRP induction and a significant decrease in fluorescence in response to N stimuli relative to F stimuli. ΔF/F calculated as F_stim – F_gray/F_gray. Replotted from Hayden et al. (2021). (D,E) Chemogenetic suppression of PV+ cells occludes SRP. (D) PV-Cre mice were injected with a virus expressing a Cre-dependent inhibitory DREADD receptor (hM4Di). The mice then underwent a standard SRP protocol, and following the familiar/novel test on day 7 received a systemic injection of CNO in order to activate hM4Di receptors and suppress PV+ cell activity. 1 h later mice viewed the original familiar stimulus and a new novel stimulus. (E) Mice showed normal SRP expression before CNO injection, but after injecting CNO to suppress PV+ cell activity mice showed an increase in VEP magnitude which eliminated the familiar-novel difference. (F,G) Optogenetic activation of PV+ cells partially reverses SRP. (F) PV-Cre mice were injected with a virus expressing Cre-dependent ChR2 and implanted with an optic fiber above bV1. Following SRP induction, mice viewed familiar and novel stimuli with blue-light activation of PV+ cells on alternating blocks on the test day. (G) Mice showed normal SRP expression during blocks in which PV+ cells were not activated by blue light, but the familiar-novel difference was significantly reduced for light-on blocks, due to a reduction in VEP magnitude. Panels (D–G) replotted from Kaplan et al. (2016). Asterisks denote significant differences throughout while non-significant differences are shown as n.s. In the final panel, actual p-values are provided.

FIGURE 4

Models of SRP. (A) Neurons in the dorsal lateral geniculate nucleus (dLGN) project to several populations of neurons in V1, including excitatory principal cells (gray triangles) and inhibitory interneurons (black circles) in multiple layers. SRP was originally interpreted as potentiation of a subset of thalamic inputs onto L4 principal cells (indicated by yellow arrowhead and + symbol), which would lead to an increased response of L4 principal cells (indicated by yellow halo) for familiar stimuli. The bulk of the evidence argues against this simple model of SRP. (B) Following SRP induction, somatostatin-positive (SOM+) inhibitory interneurons show heightened activity for familiar stimuli, and parvalbumin-positive (PV+) inhibitory interneurons show reduced activity for familiar stimuli (indicated by blue halo). This could be mediated by the potentiation of thalamic inputs onto L4 SOM+ cells, which inhibit L4 PV+ cells. (C) V1 receives a great deal of top-down feedback from other cortical and subcortical areas, including V2, retrosplenial cortex (RSP), and the basal forebrain. Changes in the feedback from these brain regions might explain some features of SRP, though how this model would also incorporate the necessity of local plasticity in V1 is unclear. (D) L6 corticothalamic cells also receive substantial thalamic input and have been implicated in SRP. Potentiation of thalamic inputs onto L6 principal cells might alter L4 activity via a population of PV+ cells known to inhibit other cortical layers. (E) Alternatively, L6 corticothalamic neurons could shift neurons in the dLGN from a tonic firing to a burst firing mode, via feedback to the thalamic reticular nucleus (TRN). This change in firing patterns could then preferentially recruit L4 SOM+ cells over PV+ cells to alter ongoing activity in L4.