Primary somatosensory cortex bidirectionally modulates sensory gain and nociceptive behavior in a layer-specific manner

Fig. 1. Layer 6 corticothalamic (L6-CT) activation in the S1 hindlimb cortex (S1HL) elicits nocifensive behaviors in the absence of peripheral stimulation. a ChR2-EYFP-expression (green) in S1HL of an L6-ChR2 mouse showing fluorescence in L6-CT neurons and their axons in the ventral posterolateral thalamus (VPL) and thalamic reticular nucleus (TRN). S1HL cortex (right panel) with L6-CT neurons depth-registered relative to S1HL layer borders (dashed lines, estimated based on soma sizes and densities using DAPI signals, blue). A representative example from n = 21 mice. b–d S1HL cortex silicon-probe recording from L6-ChR2 mice demonstrating optogenetic control of L6-CT units. b L6-CT units were identified based on their short-latency, low-jitter response to 10 ms light pulses (Supplementary Fig. 2). Each marker shows unit depth vs. mean ± SD latency to first evoked spike (pooled from 3 animals, laser strength: 1201 mW/mm²). c Example rasters for a laser-responsive L6-CT unit (depth = 1205 μm). d Blue bars: fraction of laser-responsive L6 units (from n = 232 L6 units total) as a function of laser power. Black: corresponding mean spiking rate (mean ± SEM, n = 92 units; mean depth ± SD = 1194…

Fig. 2. Layer 6 corticothalamic (L6-CT) activation in the S1 hindlimb cortex (S1HL) increases mechanical sensitivity, exacerbates inflammatory allodynia, and induces aversion. a Schematic of von Frey setup to quantify mechanical sensitivity in response to graded force stimulation of the hindpaw with and without optogenetic stimulation of L6-CT in contralateral S1HL. b Within-animal comparison of paw withdrawal probabilities in response to graded von Frey stimulation of the left hindpaw at baseline (black, Laser off) and during L6-CT laser stimulation (blue, Laser on, 5 s continuous pulse) stimulation in the contralateral S1HL of L6-ChR2 mice (n = 10, p < 0.001). L6-EGFP control animals in Supplementary Fig. 3b. c Same as in (b) but after animals were injected with Complete Freund’s adjuvant (CFA) in the left hindpaw to induce paw inflammation (n = 10 mice). For comparison of pre- and post-CFA withdrawal probabilities, see Supplementary Fig. 5. d Comparison of filament forces that evoked paw withdrawal in 60% of the stimulation trials as a function of L6-CT activation without (Naive, p < 0.0001) and with (CFA) inflammation (n = 10 mice). Lower withdrawal thresholds indicate increased…

Fig. 3. Layer 6 corticothalamic (L6-CT) activation enhances ventral posterolateral thalamus (VPL) spiking output. a Left: Schematic of VPL silicon-probe recordings in anesthetized mice (n = 4). Right: stimulation protocol for three interleaved stimulation conditions: L6-CT activation (L), mechanical paw stimulation (M), and combined L6-CT + mechanical simulation (ML). Windows for unit spike analysis is indicated in orange. b Population overlap in L/M/ML conditions. 742/1018 VPL units were significantly modulated in at least one stimulus condition; 92 units were only responsive to ML. c Proportion and counts of enhanced and suppressed VPL units for each L/M/ML condition. For example, ~40% of the VPL units were enhanced by mechanical stimulation (M), while during the ML condition, 75% were enhanced and ~10% were suppressed. A minority of units had only changes in spike timing, not spike counts (counts at zero). d Example VPL single-unit responses to L, M, and ML stimulation. Top: raster examples; middle: mechanical pressure (gray shading) and first temporal derivative of pressure (dP/dt, black); bottom: corresponding PSTHs (bin size = 20 ms, smoothed with BARS method76). e–g ML condition…

Fig. 4. Layer 6 corticothalamic (L6-CT) activation suppresses spiking in layer 5 but enhances spiking in superficial layers. a Upper: Experimental configuration for layer-resolved silicon-probe recordings in S1 hindlimb cortex (S1HL) of anesthetized mice (n = 3). Lower: stimulation protocol for three interleaved stimulation conditions: L6-CT activation (L), mechanical paw stimulation (M), and combined L6-CT + mechanical simulation (ML). Time windows for unit spike analysis are indicated in orange. b Example responses for L, M, and ML conditions in L2/3 (light gray), L4 (gray), L5 (red), and L6 (blue). Top: raster examples; middle: mechanical pressure (gray shading) and first temporal derivative of pressure (dP/dt, black); bottom: corresponding PSTHs (bin size = 20 ms, smoothed with BARS method76). c Breakdown of layer-specific population responses to L/M/ML conditions showing proportion and counts of enhanced and suppressed units for each L/M/ML condition. A minority of units had only changes in spike timing, not spike counts (counts at zero). See Supplementary Fig. 8e for the population overlap of L/M/ML—sensitive units in each layer. d L6-CT stimulation enhances superficial layers’ paw…

Fig. 5. Layer 5 (L5) inhibition in the S1 hindlimb cortex (S1HL) reproduces pronociceptive effects of Layer 6 corticothalamic (L6-CT) activation. a stGtACR2-FusionRed-expression (red) in L5 S1HL of an Rbp4-Cre mouse showing fluorescence in L5 neurons, S1HL layer borders (dashed lines), estimated based on soma sizes and densities using DAPI signals (blue). A representative example from n = 15 mice. b Schematic of silicon-probe recordings in S1HL in a head-fixed awake mouse (upper); example raster demonstrating inhibition of spiking in an example L5 unit (lower). c Example raster for a laser-suppressed L5 stGtACR2 unit (depth = 589 μm). d Upper panel: depth-resolved light modulation indices (MIL) for L2/3, L4, L5, and L6 units in S1HL during optogenetic inhibition of L5 (n = 3 pooled experiments, 285 units). Each unit is represented by one data point; the size and color of markers are proportional to the spontaneous spiking rate (black to yellow: low to high). Depth values are slightly jittered for visibility. Non-significantly modulated units are not shown (122/285). Lower panel: MIL distributions for significantly modulated units by layer. e Within-animal comparison of paw withdrawal…

Fig. 6. Layer 5 (L5) activation in the S1 hindlimb cortex (S1HL) is antinociceptive. a ChR2-EYFP-expression (green) in S1HL of a L5-ChR2 mouse showing fluorescence in L5 neurons. S1HL cortex (right panel) with L5 neurons depth-registered relative to S1HL layer borders (dashed lines, estimated based on soma sizes and densities using DAPI signals, blue). A representative example from n = 19 mice. b, c Representative S1HL cortex silicon-probe recording from an L5-ChR2 mouse demonstrating optogenetic control of the L5 pathway. b L5 units were identified based on their short-latency, low-jitter response to 10 ms light pulses (1201 mW/mm²). Each marker shows unit depth vs. mean ± SD latency to the first evoked spike (n = 32 out of 203 units recorded at a depth of L5, data from n = 2 mice). c Example raster plot of one unit from b in response to 5 s laser stimulation (1201 mW/mm2). d Optogenetically-evoked pain-like behaviors (i.e., paw shaking and withdrawal) were absent in the case of L5 stimulation (red, n = 6). L6-CT (blue, n = 11, replotted from Fig. 1f). Behavioral responses were considered in the 5 s optogenetic stimulation period, paw lifting (solid line), and limb shaking (dashed…

Fig. 7. Bidirectional modulation of sensory gain and nociception via S1 Layer 6 corticothalamic (L6-CT) and Layer 5 (L5). a Withdrawal probabilities during L5 (red, n = 6 mice) or L6-CT (blue, n = 10 mice) optogenetic stimulation (5 s continuous laser above S1 hindlimb cortex (S1HL)) as a function of filament force, expressed as changes from baseline measurements (stippled line, Laser off trials). p < 0.001. b Comparison of mechanical sensitivity, computed from areas under the curves (AUC) from naive and Complete Freund’s adjuvant (CFA)-inflamed mice, with (Laser on) and without (Laser off) activation of L5 and L6-CT neurons in S1HL. Increased AUCs correspond to increased sensitivity (hypersensitivity) and decreased AUCs correspond to decreased sensitivity (hyposensitivity). Numbers of mice: Naive L5 n = 6, p < 0.001; Naive L6-CT n = 10, p < 0.001; CFA L5 n = 12, p < 0.001; CFA L6-CT n = 10, p = 0.84. c Comparison of filament forces that evoked paw withdrawal in 60% of the stimulation trials measured in naive and CFA-inflamed mice with (Laser on) and without (Laser off) activation of L5 and L6-CT neurons in S1HL. Data points below the diagonal (dashed line) correspond to increased…

Fig. 8. Summary of bidirectional modulation of sensory gain and nociceptive behavior by S1 hindlimb cortex (S1HL) layer 5 and Layer 6 corticothalamic (L6-CT) neurons. Main findings of this study in the context of known thalamocortical circuitry of the S1 cortex6, 9, 20. L6-CT activity increases sensory gain and induces nociception via enhancement of thalamocortical transmission through L6-CT glutamatergic projections to ventral posterolateral thalamus (VPL) and feedforward inhibition of L5 neurons.