Movement Initiation Signals in Mouse Whisker Motor Cortex

Abstract

Frontal cortex plays a central role in the control of voluntary movements, which are typically guided by sensory input. Here, we investigate the function of mouse whisker primary motor cortex (wM1), a frontal region defined by dense innervation from whisker primary somatosensory cortex (wS1). Optogenetic stimulation of wM1 evokes rhythmic whisker protraction (whisking), whereas optogenetic inactivation of wM1 suppresses initiation of whisking. Whole-cell membrane potential recordings and silicon probe recordings of action potentials reveal layer-specific neuronal activity in wM1 at movement initiation, and encoding of fast and slow parameters of movements during whisking. Interestingly, optogenetic inactivation of wS1 caused hyperpolarization and reduced firing in wM1, together with reduced whisking. Optogenetic stimulation of wS1 drove activity in wM1 with complex dynamics, as well as evoking long-latency, wM1-dependent whisking. Our results advance understanding of a well-defined frontal region and point to an important role for sensory input in controlling motor cortex.

Keywords: action potential; membrane potential; motor coding; motor cortex; movement initiation; multisite silicon probe recording; optogenetics; sensorimotor integration; whisker motor control; whole-cell recording.

Figure 1

wM1 Plays a Causal Role in Initiation of Exploratory Whisking (A) AAV encoding tdTomato was injected into whisker primary somatosensory cortex (wS1) (left). Serial coronal sections reveal the wS1 pattern of innervation in frontal cortex (right). (B) Example (top left) and grand average (top right) normalized fluorescence intensity map of the wS1 axons in frontal cortex. Contour plots at half-maximum of the normalized fluorescence intensity for four mice (bottom left) show the location of wS1 axons in frontal cortex. Average normalized fluorescence intensity (n = 4 mice) plots across the anterio-posterior and medio-lateral axes (bottom right) show that the wS1 axons peak around 1 mm anterior and 1 mm lateral with respect to bregma. (C) Widefield image of a fixed brain where a conditional ChR2-expressing virus was injected into wM1 (top left). Coronal section showing the injection site localized to wM1 (top right). Three example traces (green) and average trace (black) of the whisker position upon 50 Hz blue light stimulation (bottom left). Grand average trace of the whisker position (black) for six mice upon 50 Hz blue light stimulation of wM1. Only trials without whisking in the prestimulus period were analyzed. Power spectral density of the wM1-driven whisker movement (bottom middle). Green traces are from individual mice and the black trace is the grand average spectrum. The probability of initiating whisker movements, P(Whisk), upon wM1 stimulation is high and the average whisker angle is positive, indicating a protraction (bottom right). Green circles indicate individual mice. Black circle indicates the mean. Boxplots indicate median and interquartile range. (D) Inactivation of wM1 was carried out in VGAT-ChR2 mice (top left). Widefield image showing the surface vasculature and the bone over wM1 (dotted yellow circle) that was thinned prior to inactivation (top right). Bregma (blue circle) and the lateral and midline sutures (blue dotted lines) are also shown. Only trials without whisking in the prestimulus period were analyzed. Three example whisker traces (green) during Catch trials and during wM1 opto-inactivation (bottom left). Note the increased number of failures to initiate whisking during wM1 inactivation. Quantified across animals, the probability to initiate whisking, P(Whisk), was significantly smaller during wM1 inactivation trials compared to Catch trials (bottom right). Gray lines indicate individual mice and black circles indicate mean. Boxplots indicate median and interquartile range. See also Figure S1 and Table S1.

Figure 2

Membrane Potential and AP Dynamics in wM1 during Whisker Movement Initiation (A) Example V_m recordings from L2/3 (red) and L5 (blue) neurons aligned to whisker movement onset (green) (top). Lighter V_m traces indicate single trials and darker traces indicate mean. Note the pronounced V_m hyperpolarization in the L2/3 neuron near movement onset and the V_m depolarization in the L5 neuron before movement onset. Grand average V_m traces for L2/3 (red) and L5 (blue) aligned to whisker movement onset (middle). Four epochs of interest are delineated (B, Baseline; P, Pre-movement; M, Movement-onset; L, Late during ongoing whisking). Changes in membrane potential (ΔV_m) quantified across the different epochs (bottom). On average, L2/3 neurons hyperpolarized significantly during Movement-onset, but depolarized significantly during the Late period. L5 neurons depolarized significantly during the Pre-movement period and remained depolarized during Movement-onset and Late periods. Circles indicate mean. Boxplots indicate median and interquartile range. (B) Raster plots and corresponding peri-stimulus time histograms (PSTHs) for L2/3 (red) and L5 (blue) units, aligned to whisker movement onset (green) (top). Grand average AP rates for L2/3 and L5 aligned to whisker movement onset (middle). Changes in AP rate (ΔAP rate) quantified across the different epochs (bottom). On average, L2/3 units significantly reduced AP firing rates during the Movement-onset and Late periods. L5 units significantly increased AP firing rates during the Pre-movement and Movement-onset periods but returned to Baseline during the Late period. Circles indicate mean. Boxplots indicate median and interquartile range. (C) Laminar map of spiking activity (top). The z-scored PSTHs of individual units (100 ms bin size) were aligned to whisking onset and sorted according to their depth. A smoothing window (with size of 5 units) was applied across depth to obtain the smooth activity map. Note the distinct activity patterns in L2/3 and L5. Percentage of wM1 units with significant changes in AP rate for L2/3 (middle) and L5 (bottom) across different epochs. Blue and yellow coloring indicates significant decrease and increase in AP rates, respectively. See also Figure S2 and Table S2.

Figure 3

Fast and Slow Whisking Variables Are Encoded in V_m and AP Firing of wM1 Neurons (A) Example V_m trace (red) from an L2/3 wM1 neuron during whisking (green) (far left). Note the V_m modulation coupled to phase of whisk cycle. This V_m modulation is also evident in the protraction-triggered average (middle left). Polar plot showing magnitude of V_m modulation versus the most depolarized phase in the whisk cycle (middle right). Only cells with significant modulation are indicated for L2/3 (red) and L5 (blue). Percentage of cells with significant V_m phase modulation (far right). (B) Example protraction-triggered raster plot of an L5 unit in wM1 (far left). Each row represents a whisk cycle. The inset shows the mapping from phase to percentile. Tuning curve for the unit shown in the left panel (middle left). Note the increase in AP rate during retraction. Polar plot showing magnitude of AP rate modulation versus maximal-firing phase in the whisk cycle for units with significant modulation in L2/3 (red) and L5 (blue) (middle right). Percentage of units with significant AP rate phase modulation (far right). (C) Example V_m trace (red) from an L2/3 wM1 neuron during whisking (far left). Note V_m hyperpolarization when whisking shifts to a more protracted position. Scatterplot showing mean V_m versus whisking midpoint (middle left). Histogram of slopes for cells with significant V_m midpoint tuning in L2/3 (red) and L5 (blue) (middle right). Percentage of cells with significant V_m midpoint tuning (far right). (D) Example protraction-triggered raster plot of an L5 unit in wM1 sorted by increasing values of midpoint (far left). The inset shows the mapping from midpoint to percentiles. Tuning curve for the unit shown in the left panel (middle left). Note the higher AP rate for smaller whisk midpoints. Histogram of the distribution of slopes for units with significant monotonic AP midpoint tuning (middle right). Only L5 units showed monotonic midpoint tuning. Percentage of units with significant AP rate midpoint tuning (far right). (E) Example V_m trace (blue) from an L5 wM1 neuron during whisking (far left). Note V_m hyperpolarization when whisking amplitude increases. Scatterplot showing mean V_m versus amplitude of whisking for the example cell (middle left). Histogram of distribution of slopes for cells with significant V_m amplitude tuning in L2/3 (red) and L5 (blue) (middle right). Percentage of cells with significant V_m amplitude tuning (far right). (F) Example protraction-triggered raster plot of an L5 unit in wM1 sorted by increasing values of amplitude (far left). The inset shows the mapping from amplitude to percentiles. Tuning curve for the example unit (middle left). Note the higher AP rate for larger whisk amplitudes. Histogram of the distribution of slopes for units with significant monotonic AP amplitude tuning (middle right). Only L5 units showed monotonic amplitude tuning. Percentage of units with significant AP rate amplitude tuning (far right). See also Figure S3 and Table S3.

Figure 4

wS1 Inactivation Decreases the Probability of Initiating Whisking, Hyperpolarizes V_m in wM1, and Reduces AP Rates in wM1 (A) V_m and silicon probe recordings were carried out in wM1 while wS1 was inactivated (left). Four example whisker traces during wS1 inactivation in a VGAT-ChR2 mouse (purple) and control light application in a GAD67-GFP mouse (green) (middle). Only trials without prestimulus whisking were included in the analysis. Quantified across animals, the probability to initiate whisking, P(Whisk), was significantly smaller upon wS1 inactivation compared to control light application (right). Each colored circle corresponds to data from one mouse. Black filled circles show mean. Boxplots indicate median and interquartile range. (B) Example V_m traces from L2/3 (red) and L5 (blue) wM1 neurons (upper left). Note rapid hyperpolarization upon wS1 inactivation. Light traces indicate individual trials and dark traces indicate the average. The grand average V_m (lower left). Example raster plot and PSTH for an L5 wM1 unit (upper middle). Note drop in AP rates when wS1 is inactivated. The grand average PSTHs (lower middle). wS1 inactivation led to robust hyperpolarization and decreased AP firing in wM1 (right). Filled circles show mean. Boxplots indicate median and interquartile range. See also Figure S4 and Table S4.

Figure 5

wS1 Activation Generates a Triphasic Response in wM1 Leading to Initiation of Whisking (A) V_m and silicon probe recordings were carried out in wM1 while wS1 was optogenetically excited with a 1 ms blue light pulse (left). wS1 stimulation led to whisker movement initiation with long latencies (right). (B) Example V_m traces (blue) from an L5 wM1 neuron upon wS1 activation (left). Note the triphasic V_m response with whisker movement initiation (green) following rebound depolarization. Lighter traces indicate individual trials while dark trace indicates average across trials for that cell. Grand average V_m response (right). (C) Example raster plot and PSTH (red) for an L2/3 wM1 unit upon wS1 stimulation (left). Grand average PSTHs (right). Note triphasic AP response and initiation of whisking (green) following third phase. (D) Quantification of the change in V_m and AP rate with respect to baseline during Early (left), Inhibition (middle), and Rebound (right) phases. Filled circles show mean. Boxplots indicate median and interquartile range. See also Figure S5 and Table S5.

Figure 6

Whisking Evoked by wS1 Stimulation Depends upon wM1 and Correlates with Activity in Specific Subsets of L5 Neurons (A) Example raster plot and PSTH for an L5 wM1 unit upon 1 ms optogenetic stimulation of wS1. The trials are grouped depending on whether the stimulus initiated whisking (Whisk, green) or not (No Whisk, magenta) (upper left). Note increase in AP rate during Rebound on Whisk trials, but not on No Whisk trials. The color-coded normalized z-scored AP difference between Whisk and No Whisk trials for the example unit, together with average whisker traces (lower left). Z score activity map (Whisk – No Whisk) for all wM1 units (right). Note prominent positive AP rate difference in L5. (B) Scatterplot of AP modulation index during wS1-evoked whisking versus self-initiated whisking; each circle represents a single unit. The modulation indices did not correlate in L2/3 (top) but positively correlated in L5 (bottom), indicating that L5 neurons that are modulated during wS1-evoked whisking also tend to be similarly modulated during self-initiated whisking. (C) ChR2 (green) was expressed in wS1, and muscimol (red) was injected into wM1 (left). Example whisker traces (green) upon wS1 stimulation before and after muscimol inactivation of wM1 (middle). Quantified across animals, muscimol inactivation of wM1 significantly reduced the probability of initiating whisking upon wS1 activation (upper right). Injection of Ringer’s solution in wM1 did not affect initiation of whisking (lower right). Gray lines indicate individual mice and black circles indicate mean. Boxplots indicate median and interquartile range. (D) Schematic drawing of the wS1→wM1 sensorimotor circuit. wM1 initiates rhythmic whisking by issuing a motor command to brainstem circuitry (Rt, reticular formation; FN, facial nucleus). wS1 in turn provides tonic excitatory drive to wM1 and can trigger wM1 activation, thereby initiating rhythmic whisking. See also Figure S6 and Table S6.