An Increase in Dendritic Plateau Potentials Is Associated with Experience-Dependent Cortical Map Reorganization
Total Page:16
File Type:pdf, Size:1020Kb
An increase in dendritic plateau potentials is associated with experience-dependent cortical map reorganization Stéphane Pagèsa,1, Nicolas Chenouardb,1, Ronan Chéreaua, Vladimir Kouskoffb, Frédéric Gambinob,2,3, and Anthony Holtmaata,2,3 aDepartment of Basic Neurosciences and the Center for Neuroscience, Centre Médical Universitaire (CMU), University of Geneva, 1211 Geneva, Switzerland; and bUniversity of Bordeaux, CNRS, Interdisciplinary Institute for Neuroscience, UMR 5297, F-33000 Bordeaux, France Edited by Roberto Malinow, University of California San Diego, La Jolla, CA, and approved January 15, 2021 (received for review December 11, 2020) The organization of sensory maps in the cerebral cortex depends the thalamus (22). The paralemniscal pathway provides a com- on experience, which drives homeostatic and long-term synaptic plementary and nonoverlapping source of inputs mainly termi- plasticity of cortico-cortical circuits. In the mouse primary somato- nating in L5a and L1 that arise from the higher-order posterior sensory cortex (S1) afferents from the higher-order, posterior me- medial thalamic nucleus (POm) of the thalamus. While the VPM dial thalamic nucleus (POm) gate synaptic plasticity in layer (L) 2/3 is viewed as the main hub for whisker tactile information to S1, pyramidal neurons via disinhibition and the production of den- the exact function of the POm in this cortical area remains un- dritic plateau potentials. Here we address whether these thalamo- clear (24, 25). Neurons in the POm have broad receptive fields cortically mediated responses play a role in whisker map plasticity in (26, 27), and their axons have extensive arborizations in S1, S1. We find that trimming all but two whiskers causes a partial fusion distributed over multiple barrel-related columns (22, 28, 29). of the representations of the two spared whiskers, concomitantly with POm axons connect to distal pyramidal cell dendrites as well as an increase in the occurrence of POm-driven N-methyl-D-aspartate to various interneurons (20, 30–35). The large extent of their receptor-dependent plateau potentials. Blocking the plateau po- projections together with their broad receptive fields suggests tentials restores the archetypical organization of the sensory map. that POm neurons provide more global input to S1 as compared Our results reveal a mechanism for experience-dependent cortical to VPM neurons. NEUROSCIENCE map plasticity in which higher-order thalamocortically mediated POm projections to S1 mediate whisker-evoked N-methyl- plateau potentials facilitate the fusion of normally segregated cortical D-aspartate receptor (NMDAR)-dependent plateau potentials representations. and facilitate whisker-evoked LTP in L2/3 pyramidal neurons (36), which may depend on a combined excitation and disinhi- synaptic plasticity | dendritic signaling | posterior medial complex of the bition (34). In addition, POm projections themselves display thalamus | somatosensory | barrel cortex plasticity during sensory learning (20). Altogether, this suggests that POm projections to S1 could play a distinctive role in the ensory cortices contain functional topographic maps, which Scan rapidly change in response to training and altered sen- Significance sory experience (1). For example, whisker trimming in rodents modifies the proportional representation of spared and trimmed Here we describe a mechanism for cortical map plasticity. whiskers in the barrel field of the primary sensory cortex (S1) – Classically, representational map changes are thought to be (1 4). This type of cortical map plasticity is thought to be driven driven by changes within cortico-cortical circuits, e.g., Hebbian by long-term potentiation (LTP) and long-term depression plasticity of synaptic circuits that lost vs. maintained an excit- (LTD) of layer (L) 4-to-L2/3 and L2/3-L2/3 cortico-cortical (CC) – atory drive from the first-order thalamus, possibly steered by synapses (2, 5 7), as well as by changes in intrinsic neuronal neuromodulatory forces from deep brain regions. Our work properties and homeostatic mechanisms balancing the loss of provides evidence for an additional gating mechanism, pro- surrounding sensory inputs (8, 9). In addition, whisker trimming vided by plateau potentials, which are driven by higher-order weakens feed-forward inhibition of L2/3 pyramidal neurons thalamic feedback. Higher-order thalamic neurons are charac- (10–12), and, similarly to monocular deprivation, even may – terized by broad receptive fields, and the plateau potentials evoke pruning of inhibitory synapses (13 15). Disinhibition that they evoke strongly facilitate long-term potentiation and could also serve a role in homeostasis by increasing whisker- elicit spikes. We show that these features combined constitute evoked neuronal spiking (9) and gate synaptic plasticity (12). a powerful driving force for the fusion or expansion of sensory Thalamo-cortical (TC) synapses may play a direct or facili- representations within cortical maps. tating role in cortical map plasticity. TC axons have been shown to remain plastic throughout life and to be affected by modifi- Author Contributions: S.P., N.C., R.C., V.K., and F.G. performed research and analyzed cations of sensory experience (16–19). Trimming a subset of data; and F.G. and A.H. conceived the studies, supervised the research, and wrote whiskers causes a decrease in TC innervation of deprived but not the paper. of spared barrels (16, 18), and sensory learning may induce The authors declare no competing interest. plasticity of a subset of TC synapses (20). However, the relative This article is a PNAS Direct Submission. contributions of CC and TC synaptic plasticities, and how they This open access article is distributed under Creative Commons Attribution-NonCommercial- interact during cortical map plasticity, is not known. Moreover, NoDerivatives License 4.0 (CC BY-NC-ND). the role of TC synapses may be intricate since different cortical 1S.P. and N.C. contributed equally to this work. layers receive inputs from diverse thalamic origins, each with 2F.G. and A.H. contributed equally to this work. distinctive properties. 3To whom correspondence may be addressed. Email: [email protected] or Sensory information from the whiskers is transmitted to S1 by [email protected]. two main and well-segregated TC projections (21–23). The This article contains supporting information online at https://www.pnas.org/lookup/suppl/ lemniscal pathway relays sensory information to L5b, L4, and L3 doi:10.1073/pnas.2024920118/-/DCSupplemental. neurons through the ventral posteromedial (VPM) nucleus of Published February 22, 2021. PNAS 2021 Vol. 118 No. 9 e2024920118 https://doi.org/10.1073/pnas.2024920118 | 1of8 Downloaded by guest on September 27, 2021 refinement of cortical maps. Here, we investigate the relation- 48 h A 24 h ship between cortical remapping and POm-mediated plateau 5 min potentials upon whisker sensory deprivation. We use a paradigm in which all whiskers were trimmed except a pair of neighboring 1 s / 8 Hz 1 s / 8 Hz 1 s / 8 Hz ones (dual-whisker experience, DWE). Using intrinsic optical ccd ccd ccd imaging, we first corroborate previous electrophysiology studies which showed that DWE causes the representation of the two spared whiskers to partly fuse (37, 38). We then show that this naive FWE (DWE 0) DWE 1 DWE 2 plasticity is associated with an increase in dendritic plateau po- B baseline 5 tentials, which is dependent on inputs from the POm. The t-test pharmacological removal of the plateau potentials causes the -2 WRD value fused whisker representations to segregate back to an organiza- stim. t tion seen in naive mice. Altogether, our results reveal a mech- 0.2 mm anism for rapid experience-dependent cortical map plasticity, -12 t map low-pass filtered thresholded which consists of an increased contribution of dendritic plateau potentials that are associated with inputs from higher-order * C *** ** D p<0.001 thalamic neurons. This, in turn, may enhance the level of non- 350 ns 1.0 350 *** specific sensory input and facilitate subsequent synaptic plasticity events that have been shown to underlie cortical map reorgani- 300 300 zation (2–4, 6). 250 250 0.5 200 200 Results WRD (µm) probability WRD (µm) 150 150 Dual-Whisker Experience Reshapes Whisker-Evoked Intrinsic Optical FWE (DWE 0) DWE 2-4 100 31 21 Signals in S1. Single-unit and whole-cell recordings have shown 100 319 8 9 4 0.0 100 350 01234 0 that DWE causes the functional representation of the spared 2-4 whiskers in S1 to merge (4, 12, 37, 38). Intrinsic optical signal DWE WRD (µm) DWEDWE (IOS) imaging, which is a proxy of whisker-evoked population activity (39, 40), can potentially quantify such changes at the Fig. 1. IOS detects DWE-evoked plasticity of whisker representation in S1. mesoscale level in a quasi-noninvasive manner (41–43). Here, we (A) Schematic of whisker trimming and IOS recording. (B, Left) Examples of used IOS imaging to measure DWE-evoked plasticity of whisker averaged baseline and stimulus-related raw IOS evoked by one train of representations in S1. whisker deflections. (Middle) unfiltered and low-pass-filtered t-value maps − Mice were separated in two groups. One group was exposed to computed from 20 stimulations. Only pixels with a t-value lower than 2 are a brief period of DWE (1 to 4 d) by clipping all whiskers except included in the responding area (red dashed line). (Right) Whisker C1 (blue) and C2 (red) responding areas. WRD: Euclidian distance between the peaks whisker C1 and C2, while for the control group all whiskers were of the C1 and C2 response areas. (C) Median (± interquartile range) WRD as left intact to allow a full whisker experience (FWE). We used a function of deprivation duration (in days). (D, Left) Cumulative distribution IOS to assess the spatial representation of the C1 and C2 whis- of WRD in control mice (FWE) and upon DWE. (Right) Median (± inter- kers in S1 under urethane anesthesia (Fig.