Translaminar Cortical Membrane Potential Synchrony in Behaving Mice

Abstract

The synchronized activity of six layers of cortical neurons is critical for sensory perception and the control of voluntary behavior, but little is known about the synaptic mechanisms of cortical synchrony across layers in behaving animals. We made single and dual whole-cell recordings from the primary somatosensory forepaw cortex in awake mice and show that L2/3 and L5 excitatory neurons have layer-specific intrinsic properties and membrane potential dynamics that shape laminar-specific firing rates and subthreshold synchrony. First, while sensory and movement-evoked synaptic input was tightly correlated across layers, spontaneous action potentials and slow spontaneous subthreshold fluctuations had laminar-specific timing; second, longer duration forepaw movement was associated with a decorrelation of subthreshold activity; third, spontaneous and sensory-evoked forepaw movements were signaled more strongly by L5 than L2/3 neurons. Together, our data suggest that the degree of translaminar synchrony is dependent upon the origin (sensory, spontaneous, and movement) of the synaptic input.

Graphical abstract

Figure 1

Distinct Membrane Properties of L2/3 and L5 Cortical Pyramidal Neurons (A) Image of the glabrous skin of the right forepaw showing five digits (D1–D5), three central pads (P1–P3), the thenar pad (TH), and the hypothenar pad (HT) (Waters et al., 1995). Scale bar, 1 mm. (B) Cartoon schematic showing head-fixed awake mouse with recording electrodes in red (L2/3) and blue (L5), with forepaw digit movement (green) monitored by the sensing arm (gray) that was also used for tactile stimulation. (C) Biocytin reconstructions of L2/3 (red) and L5 (blue) neurons, with axons in lighter color, next to a histogram showing the depths of all recorded L2/3 and L5 neurons (n = 17 L2/3 neurons and n = 28 L5 neurons) based on micromanipulator reading and biocytin staining. (D) Three single trial responses of a L2/3 (red) and a L5 (blue) pyramidal neuron to intracellular current injection with different amplitudes (from top to bottom: +400 pA, +300 pA, and +200 pA). The L2/3 example corresponds to the reconstructed L2/3 neuron in (C). Horizontal lines indicate −60 mV for L2/3 and L5. (E) Plotting the evoked spike rate as a function of current injection amplitude reveals that L5 neurons are more excitable than L2/3 neurons. Filled circles with error bars show mean ± SEM. (F) L5 neurons have a significantly larger input resistance than L2/3 neurons during hyperpolarizing current injection. Open circles show individual cells. (G) L5 neurons show a significantly larger amplitude hyperpolarization following positive current injection (afterhyperpolarization [AHP]) than L2/3 neurons at all current amplitudes tested. For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001.

Figure 2

Laminar-Specific V_m Dynamics of L2/3 and L5 Neurons during Forepaw Behavior (A) Example whole-cell recordings from a L2/3 neuron (red) and L5 neuron (blue) with the digit movement (green) measured by the stimulator/sensing arm in contact with the glabrous skin of forepaw digit 3. (B) Population average fast Fourier transforms (FFTs) of the V_m of L2/3 and L5 neurons during quiet wakefulness (black) and digit movement (orange). (C) Power of low-frequency activity (1–5 Hz) in the quiet (Q) and moving (M) state shows a significant reduction during digit movement in both L2/3 and L5 neurons. Filled circles with error bars show mean ± SEM, and lines represent individual neurons. (D) SD of the V_m was significantly reduced during forepaw movement in both L2/3 and L5 neurons. (E) L5 neurons showed an overall higher AP firing rate than L2/3 neurons in quiet and moving mice and a significant increase in AP firing rate during movement. (F) AP threshold was not significantly different at rest and during movement in L2/3 and L5 neurons. (G) Both L2/3 and L5 neurons depolarize during digit movement, but L2/3 neurons are more hyperpolarized than L5 neurons in both behavioral states. (H) The mean firing rate (Q and M periods) of L2/3 and L5 neurons plotted as a function of the distance between AP threshold and the mean value of the maximum 10% of the V_m (Max V_m). Filled circles show the mean value for one cell. For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. See also Figures S1–S3.

Figure 3

V_m Synchrony between L2/3 and L5 Neurons Is Dependent on Behavioral State (A) Example dual whole-cell recording from a L2/3 (red) and a L5 (blue) cortical neuron during digit movement (green). (B) Mean cross-correlation for the example recording shown in (A), taking L5 as the reference, shows a higher correlation during quiet (Q) than digit movement (M). (C) Example mean coherence spectrum from the example recording shown in (A) from Q and M periods. (D) Population mean cross-correlation (n = 9 pairs) during Q and M periods. (E) Significant reduction in the peak cross-correlation value in M compared to Q periods. Filled circles with error bars show mean ± SEM, lines show data from individual pairs. (F) The peak time of the cross-correlation shows a positive lag indicating that L5 neurons are active before L2/3 neurons in Q and M periods. (G) Population mean average of coherence spectrum during Q and M periods (n = 8 pairs). (H) A significant reduction in coherence from Q to M periods in frequency band 1–5 Hz. (I) Population V_m average of L2/3 (red) and L5 neurons (blue) centered on APs in L2/3 neurons during quiet (left, n = 8 pairs) and moving (right, n = 4 pairs) periods. Bottom, corresponding population L5 spike-time PSTHs. (J) Same as (I) but for L5 spike-triggered averages with L2/3 PSTH below (n = 8 pairs). (K) Quantification of the V_m rise time in L2/3 and L5 neurons between −22 ms and −2 ms before a (left) L2/3 AP and (right) L5 AP in quiet and moving periods. Filled circles show population mean with error bars showing mean ± SEM. Gray lines show values from individual cells. For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001.

Figure 4

Slow Depolarizing Fluctuations during Quiet Wakefulness Have an Earlier Onset in L5 Neurons (A) Example slow depolarizing events (SDEs) from a dual L2/3 (red) and L5 (blue) whole-cell recording in an awake, resting mouse. V_m traces are aligned to the onset (left) and offset (right) of the SDE in the L2/3 neuron, and bottom traces show V_m averages. Note that L5 leads at the onset, but not the offset, of the SDEs. Horizontal lines indicate V_m (mV) for L2/3 and L5 (trial 1 onset, −69.3/−63.1 mV and offset, −47.7/−44.7 mV; trial 2 onset, −59.3/−59.1 mV and offset, −40.0/−37.7 mV; trial 3 onset, −60.9/−57.0 mV and offset, −31.8/−31.7 mV; trial 4 onset, −59.7/−56.4 mV and offset, −42.0/−37.7 mV; and trial 5 onset, −58.3/−55.7 mV and offset, −40.0/−30.1 mV). Average onset, −60.0/−52.6 mV and offset, −41.9/−37.0 mV. APs have been truncated. (B) Plots of selected SDEs from seven dual whole-cell recordings. SDEs were aligned at threshold crossing at the onset (left) and offset (right) of the SDE in the L2/3 neuron and arranged by duration. Top boxes show L2/3 data, and bottom boxes show L5 with colors corresponding to the normalized V_m from minimum (blue) to maximum (red) values. (C) Population distribution (top) and trial-by-trial measurements (bottom) of the subthreshold onset (left) and offset (right) times in L5 neurons relative to the onset and offset times in L2/3, respectively (n = 7 dual recordings). Onset and offset times were estimated by the 5% level of a sigmoidal fit to the V_m at onset and offset (see Supplemental Experimental Procedures for details). (D) Population average of the normalized V_m SDEs in L2/3 and L5 relative to the threshold-crossing at onset and offset of the L2/3 neuron. (E) Population peri-SDE time histogram of AP times from the dataset used in (D). (F) Population analysis of onset and offset times relative to the L2/3 SDE shows a significantly earlier onset in L5 but similar offset times. To calculate the onset/offset timing difference, we first measured the time of SDE onset/offset relative to the time of threshold crossing of the L2/3 SDE. Then, we subtracted the population mean L2/3 onset/offset time from all values. Filled circles with error bars show mean ± SEM, and lines show data from individual pairs. For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. See also Figures S4 and S5.

Figure 5

Movement Onset Synchronizes Synaptic Input across Layers and Results in an Increase in L5 AP Rate (A) Population V_m average responses of L2/3 (red) and L5 (blue) neurons to onset of spontaneous digit movement (green shows the rectified first derivative of the digit movement (digit_FD); n = 11 L2/3 cells and n = 19 L5 cells, from single and dual recording experiments). Movement onsets (dashed vertical line) were detected via thresholding of digit_FD (see Experimental Procedures). (B) Peak cross-correlation between the digit_FD and the V_m shows no significant time lag between layers, indicating synchronous depolarization. Filled circles with error bars show mean ± SEM, and open circles show individual cells. (C) Mean population AP rates over time with respect to movement onset from L2/3 and L5 neurons. (D) L5 neurons show a significantly higher AP firing rate after movement onset (between 0 and 1 s) than L2/3 neurons. (E) Population average of the variance of the V_m over time with respect to movement; gray sections and numbers show time points for analysis in (F). (F) L2/3 and L5 neurons show a significant reduction in the V_m variance following movement onset; the two measurement time windows (1 and 2) are indicated in (E). For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001.

Figure 6

Tactile-Evoked Subthreshold Responses Are Highly Correlated across Cortical Layers and Modulated by Behavioral State (A) Single-trial tactile-evoked responses from a dual L2/3 (red) and L5 (blue) whole-cell recording during quiet wakefulness (left) and movement (right). (B) Top: mean V_m tactile-evoked response of example neurons in (A) during quiet wakefulness and movement shows reduction in amplitude during moving periods. Bottom: corresponding PSTH of AP firing. (C) Population averaged subthreshold tactile-evoked response during quiet wakefulness and movement. (D) Population PSTH from L2/3 and L5 to tactile stimulation during quiet wakefulness and movement. (E) L2/3 and L5 neurons show a decrease in tactile-evoked subthreshold response amplitude as mice go from quiet (Q) to movement (M). Filled circles with error bars show mean ± SEM. Lines show individual cells. (F) Population analysis of background-subtracted AP firing rates to tactile stimulation of the forepaw shows no difference between Q (left) and M (right) periods in both layers. AP rate measured as the difference between the 100 ms before and 100 ms after stimulus onset. (G) The subthreshold tactile-evoked reversal potential shows no significant difference between layers or behavioral states. (H) Absolute AP rate in the 100 ms after stimulus onset plotted as a function of the difference in V_m between AP threshold and the tactile stimulus-evoked response reversal potential. Filled circles show mean value of individual cells. (I) Subthreshold tactile-evoked response latencies of L2/3 and L5 neurons are not significantly different. (J) Amplitude of L2/3 subthreshold tactile-evoked responses plotted against the amplitude of L5 subthreshold responses from the example pair in (A) shows highly correlated response amplitudes during Q (black) and M (orange) periods. (K) Population data of cross-correlation of mean subthreshold tactile-evoked responses (combining Q and M responses) between L2/3 and L5 neurons. Filled circles with error bars show mean ± SEM, and open circles show individual cells. For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. See also Figures S6 and S7.

Figure 7

Tactile-Evoked Forepaw Movements Are Signaled by L5 Neurons (A) Mean V_m tactile-evoked responses with corresponding digit movement (green) and PSTHs from an L2/3 (red, top) and an L5 (blue, bottom) neuron in resting, quiet mice that showed no behavioral response (left, quiet-quiet [QQ]) or a short-latency digit movement following the stimulus (right, quiet-movement [QM]). (B) Grand average tactile-evoked responses from all L2/3 and L5 neurons during QQ and QM trials. (C) Population PSTHs of firing rates in L2/3 and L5 neurons following tactile stimulation in QQ and QM trials. Note that only L5 neurons show an evoked spiking response in the later phase (300–400 ms after stimulus onset). (D) The amplitude of the V_m response to tactile stimulation is significantly larger for QM trials than in QQ trials in both L2/3 and L5 neurons. Filled circles with error bars show mean ± SEM. Lines show individual cells. (E) The mean V_m in the 100 ms before stimulus onset is more hyperpolarized in QM trials than QQ trials in L5, but not in L2/3 neurons. (F) The mean V_m in the late phase (300–400 ms following tactile stimulation) is significantly more depolarized after a QM trial than a QQ trial in both layers. (G) Significant increase (background subtracted) in AP firing rates in QM trials as compared to QQ trials in the late phase (300–400 ms following tactile stimulation) in L5 neurons, but not in L2/3 neurons. For all panels, ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001.