The dorsal column nuclei scale mechanical sensitivity in naive and neuropathic pain states

Abstract

During pathological conditions, tactile stimuli can aberrantly engage nociceptive pathways leading to the perception of touch as pain, known as mechanical allodynia. The brain stem dorsal column nuclei integrate tactile inputs, yet their role in mediating tactile sensitivity and allodynia remains understudied. We found that gracile nucleus (Gr) inhibitory interneurons and thalamus-projecting neurons are differentially innervated by primary afferents and spinal inputs. Functional manipulations of these distinct Gr neuronal populations bidirectionally shifted tactile sensitivity but did not affect noxious mechanical or thermal sensitivity. During neuropathic pain, Gr neurons exhibited increased sensory-evoked activity and asynchronous excitatory drive from primary afferents. Silencing Gr projection neurons or activating Gr inhibitory neurons in neuropathic mice reduced tactile hypersensitivity, and enhancing inhibition ameliorated paw-withdrawal signatures of neuropathic pain and induced conditioned place preference. These results suggest that Gr activity contributes to tactile sensitivity and affective, pain-associated phenotypes of mechanical allodynia.

Keywords: CP: Neuroscience; DCN; LTMR; PSDC; brain stem circuits; dorsal column nuclei; low-threshold mechanoreceptors; mechanical allodynia; neuropathic pain; postsynaptic dorsal column neurons; tactile perception.

Figure 1.. Ascending DRG and spinal pathways converge in the gracile nucleus

(A) The gracile nucleus (Gr, green) receives sensory inputs from lower body DRG afferents (cyan) and spinal projections (magenta). Retrograde virus injections into the VPL (right) labeled Gr VPL-projecting neurons (PNs). (B) Injection of cholera toxin β subunit (CTb, cyan) into the hindpaw glabrous skin labeled hindlimb afferents targeting the Gr (white dashed circle). Scale bar, 200 μm. (C) Advillin^Cre;Rosa26^{LSL-Synaptophysin-GFP} mice labeled synaptic terminals of primary afferents (cyan) in the Gr core (solid white line) and shell (dashed white line). Scale bar, 100 μm. (D) Cdx2^Cre;Lbx1^FlpO;Rosa26^{LSL-FSF-Synaptophysin-GFP} mice labeled terminals of spinal projections (magenta) in the Gr core (solid white line) and shell (dashed white line). Scale bar, 100 μm. (E) In situ hybridization of excitatory (vGluT2, blue) and inhibitory (vGAT, red; GlyT2, green) neurons in the Gr core (solid white line) and shell (dashed white line). Scale bar, 100 μm. (F and G) Quantification of excitatory and inhibitory neuronal markers in the Gr core (F) and shell (G). (H) Strategy to label VPL-PNs (green) or inhibitory neurons (red). Inset: representative injection of AAV-retro GFP into the VPL. Scale bar, 1,000 μm. (I and J) Representative images of Gr VPL-PNs (I) (green), inhibitory neurons (J) (red), and NeuN⁺ immunostaining of neurons (blue). Scale bar, 100 μm. (K and L) Quantification of VPL-PNs and inhibitory neurons in the Gr core (K) and shell (L) normalized to total NeuN.

Figure 2.. Gr VPL-PNs and inhibitory neurons exhibit different electrophysiological properties and excitatory drive

(A) Labeling of VPL-PNs and inhibitory neurons for slice electrophysiology. (B) Cell-attached voltage-clamp recordings showing spontaneous AP discharge in VPL-PN (green) and inhibitory neuron (red). (C) Percentage of neurons exhibiting spontaneous AP discharge. (D) Quantification of sAP frequency. (E) Whole-cell current clamp recording (−60 mV) from VPL-PN (green) and inhibitory neuron (red) in response to depolarizing current injection. (F) Quantification of AP number in response to depolarizing current injection. (G) Quantification of AP rheobase. (H) Quantification of AP threshold. (I) Whole-cell voltage-clamp recordings (−70 mV) of sEPSCs from VPL-PN (green) and inhibitory neuron (red). (J) Captured sEPSCs from representative cells shown in (I). Right trace shows amplitude normalized overlay. (K) Quantification of sEPSC frequency. (L) Quantification of sEPSC amplitude. (M) Quantification of sEPSC rise time. (N) Quantification of sEPSC decay time.

Figure 3.. Primary afferent and spinal inputs differentially innervate Gr VPL-PNs and inhibitory neurons and evoke feedforward inhibition

(A) Strategies to optogenetically activate primary afferent or spinal inputs onto Gr VPL-PNs or inhibitory neurons. (B) Whole-cell voltage-clamp recordings (−70 mV) of optically evoked EPSCs (oEPSCs) from VPL-PN (green) and inhibitory neuron (red) in response to primary afferent (top) and spinal projection (bottom) photostimulation. Ten consecutive sweeps, with average overlaid. (C) Quantification of monosynaptic oEPSC amplitude. (D) Percentage of neurons exhibiting monosynaptic oEPSCs. (E) Cell-attached voltage-clamp recording showing optically evoked AP discharge (sAP). Ten consecutive sweeps overlaid. (F) Percentage of neurons exhibiting sAP discharge. (G) Cell-attached voltage-clamp recording showing the impact of optogenetic activation of ascending inputs on spontaneous AP discharge. Ten consecutive sweeps overlaid. (H) Whole-cell voltage-clamp recordings (−20 mV) of optically evoked IPSCs (oIPSCs) in control (black) and following application of bicuculline (10 μM) and strychnine (1 μM; orange) in response to optogenetic activation of ascending inputs. Ten consecutive sweeps with average overlaid. (I) Percentage of neurons exhibiting oIPSCs. (J) Strategy to optogenetically activate inhibitory inputs onto Gr VPL-PNs. (K) Whole-cell voltage-clamp recordings (−70 mV) of oIPSCs in control (black) and following sequential application of strychnine (1 μM; blue) and bicuculline (10 μM; orange) in response to optogenetic activation of inhibitory neurons. Ten consecutive sweeps, with average overlaid. (L) Normalized impact of bicuculline (10 μM), strychnine (1 μM), and bicuculline + strychnine application on oIPSC amplitude. (M) Cell-attached voltage-clamp recording showing the impact of optogenetic activation of inhibitory inputs on spontaneous AP discharge. Ten consecutive sweeps overlaid. (N) Whole-cell voltage-clamp recordings (−70 mV) of sIPSCs from VPL-PN in control (black) and following sequential application of strychnine (1 μM; blue) and bicuculline (10 μM; orange). (O) Normalized impact of bicuculline (10 μM), strychnine (1 μM), and bicuculline + strychnine application on sIPSC frequency. (P) Normalized impact of bicuculline (10 μM), strychnine (1 μM), and bicuculline + strychnine application on sIPSC amplitude. (Q) Captured sIPSCs from representative cells shown in (N). Right trace shows amplitude normalized overlay. (R) Normalized impact of bicuculline (10 μM), strychnine (1 μM), and bicuculline + strychnine application on sIPSC decay.

Figure 4.. Manipulation of Gr VPL-PNs and inhibitory neuron scales tactile sensitivity and intensity of tactile withdrawal responses

(A and B) Strategy to silence Gr VPL-PNs (red) using CNO, with saline as a control. Scale bar, 100 μm. (C) Responses to von Frey stimulation when silencing VPL-PNs. (D) Responses to dynamic brush when silencing VPL-PNs. (E and F) Strategy to silence Gr inhibitory neurons (red) using CNO, with saline as a control. Scale bar, 100 μm. (G) Responses to von Frey stimulation when silencing Gr inhibitory neurons.*CNO vs. saline, ^#SNI vs. saline. (H) Responses to dynamic brush when silencing Gr inhibitory neurons. (I) Representative traces of MTP y-coordinates during 0.6-g von Frey stimulation for control mice (black), mice with silenced Gr inhibitory neurons (blue), and SNI mice (red). (J) Pain Assessment at Withdrawal Speeds (PAWS) analysis for SNI mice and mice where Gr inhibitory neurons were silenced compared to saline control. ↑ represents an increase compared to control mice, ↓ represents a decrease compared to control mice, and – represents no change. (K) First three dimensions of B-SOiD uniform manifold approximation and projection (UMAP) embedding, showing in low-dimensional space a projection of extracted behavioral features. Colors represent behavioral clusters identified by hierarchical density-based spatial clustering of applications with noise, including a shaking cluster (turquoise). (L) Conditioned place preference assay to assess appetitive/aversive nature of silencing Gr inhibition.

Figure 5.. Increased tactile-evoked activity and altered primary afferent drive onto Gr VPL-PNs during neuropathic pain

(A) c-Fos immunolabeling (green) of Gr neurons in response to brush stimulation in anesthetized sham and SNI mice. Scale bar, 100 μm. (B) Quantification of Fos⁺ cells (normalized to total NeuN) in response to brush stimulation in anesthetized sham and SNI mice. (C) Brush-evoked activity (gray boxes) and spontaneous activity (space between boxes) during in vivo recordings in anesthetized mice. (D and E) Quantification of evoked (D) and spontaneous (E) spiking activity. (F) Approach to activate primary afferent inputs to VPL-PNs. (G) Percentage of neurons that exhibited an optically evoked AP (oAP) following primary afferent photostimulation. (H) Cell-attached voltage-clamp recording showing oAP following primary afferent photostimulation in sham (gray) and SNI (red). Ten consecutive sweeps. Right: raster plot of example recordings. (I) Quantification of first oAP latency. (J) Quantification of first oAP latency standard deviation. (K) Quantification of first oAP reliability (percentage of trials in which primary afferent photostimulation elicited an oAP in each cell). (L) Quantification of the number of oAPs evoked per stimulus (total oAP number/number of trials). (M) Whole-cell voltage-clamp recordings of oEPSC from VPL-PNs in sham (gray) and SNI (red) mice following primary afferent photostimulation. Ten consecutive sweeps, with average overlaid. Following SNI, one-third of VPL-PNs exhibited prolonged activity after primary afferent photostimulation (right). (N) Raster plot of EPSC incidence in sham (top) and SNI (bottom) VPL-PNs. (O) Quantification of oEPSC amplitude. (P) Whole-cell voltage-clamp recording (−70 mV) from VPL-PNs in sham (gray) and SNI (red) mice showing oEPSC paired-pulse response to primary afferent photostimulation. (Q) Quantification of oEPSC paired-pulse ratio.

Figure 6.. Silencing VPL-PNs or enhancing Gr inhibition specifically alleviates tactile hypersensitivity

(A and B) Strategy to silence Gr VPL-PNs (red) in SNI mice. Scale bar, 100 μm. (C) Responses to von Frey stimulation when silencing VPL-PNs. *Uninjured + saline vs. SNI + saline; ^#SNI + saline vs. SNI + CNO; ^†uninjured + saline vs. SNI + CNO. (D) Responses to dynamic brush when silencing VPL-PNs in SNI mice. (E and F) Strategy to optogenetically activate Gr inhibition in SNI and CIPN mice implanted with an optic cannula over the Gr. Blue light activated Gr inhibition, while orange light was used as a control. (G) Responses to von Frey stimulation when activating Gr inhibition in SNI mice. *Orange control vs. blue control; ^#orange control vs. blue, 20 Hz; ^†blue control vs. blue 20 Hz. (H) Responses to dynamic brush when activating Gr inhibition in SNI mice. (I) Responses to von Frey stimulation when activating Gr inhibition in CIPN mice. (J) Responses to dynamic brush when activating Gr inhibition in CIPN mice.

Figure 7.. Enhancing Gr inhibition reduces neuropathic paw-withdrawal signatures and induces conditioned place preference

(A) Representative traces of MTP y-coordinates of SNI mice responding to a 0.6-g von Frey stimulus while activating Gr inhibition. (B) PAWS-identified paw-withdrawal features of SNI mice while activating Gr inhibition. ↑ represents an increase compared to control mice, ↓ represents a decrease compared to control mice, and – represents no change. (C) Quantification of a B-SOiD-identified shaking behavior while activating Gr inhibition in SNI mice. (D) Conditioned place preference during activation of Gr inhibition in SNI mice. (E) Representative traces of MTP y-coordinates of CIPN mice responding to a 0.6-g von Frey stimulus while activating Gr inhibition. (F) PAWS-identified paw-withdrawal features of CIPN mice while activating Gr inhibition. ↑ represents an increase compared to control, ↓ represents a decrease compared to control, and - represents no change. (G) Quantification of a B-SOiD-identified shaking behavior while activating Gr inhibition in CIPN mice. (H) Conditioned place preference during activation of Gr inhibition in CIPN mice. For all figures, data are reported as mean values ± SEM. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001. For further details on genetic crosses and statistical tests, see STAR Methods.