Sensory neuron downregulation of the Kv9.1 potassium channel subunit mediates neuropathic pain following nerve injury

Abstract

Chronic neuropathic pain affects millions of individuals worldwide, is typically long-lasting, and remains poorly treated with existing therapies. Neuropathic pain arising from peripheral nerve lesions is known to be dependent on the emergence of spontaneous and evoked hyperexcitability in damaged nerves. Here, we report that the potassium channel subunit Kv9.1 is expressed in myelinated sensory neurons, but is absent from small unmyelinated neurons. Kv9.1 expression was strongly and rapidly downregulated following axotomy, with a time course that matches the development of spontaneous activity and pain hypersensitivity in animal models. Interestingly, siRNA-mediated knock-down of Kv9.1 in naive rats led to neuropathic pain behaviors. Diminished Kv9.1 function also augmented myelinated sensory neuron excitability, manifested as spontaneous firing, hyper-responsiveness to stimulation, and persistent after-discharge. Intracellular recordings from ex vivo dorsal root ganglion preparations revealed that Kv9.1 knock-down was linked to lowered firing thresholds and increased firing rates under physiologically relevant conditions of extracellular potassium accumulation during prolonged activity. Similar neurophysiological changes were detected in animals subjected to traumatic nerve injury and provide an explanation for neuropathic pain symptoms, including poorly understood conditions such as hyperpathia and paresthesias. In summary, our results demonstrate that Kv9.1 dysfunction leads to spontaneous and evoked neuronal hyperexcitability in myelinated fibers, coupled with development of neuropathic pain behaviors.

Figures

Figure 1.

Kv9.1 is selectively expressed in myelinated sensory neurons. A, Bright-field (BF; left) and polarized light (right) images from Kv9.1 ISH in lumbar DRG showing mRNA silver grains over large (arrows), but not small (arrowheads) neurons. B, Kv9.1 cell-size distribution in rat lumbar DRG (n = 4). C, A similar expression pattern was detected using a different Kv9.1 mRNA probe (left), while control ISH reactions with excess unlabeled probe resulted in only background signal (right). D, IHC for neuronal markers combined with Kv9.1 ISH in a single lumbar DRG section. Merged images (bottom, high power on right) illustrate Kv9.1 mRNA detection in NF200-positive neurons (arrows), while CGRP-positive (arrowheads) and IB4-positive (double arrowheads) neurons are devoid of Kv9.1 signal. E, IHC showing Kv9.1 protein expression in large NF200(+), but not small (some distinguished by IB4 staining) neurons of lumbar DRG. Scale bar, 50 μm.

Figure 2.

Kv9.1 is coexpressed with Kv2 subunits in myelinated sensory neurons. A, Bright-field (top) and ISH overlay (bottom) images for Kv9.1/Kv2.1 (left) and Kv9.1/Kv2.2 (right) colocalization detection on adjacent DRG sections. Arrows denote examples of the same neurons identified in both fields, expressing Kv9.1 and Kv2.1 (blue) or Kv9.1 and Kv2.2 (red) mRNAs. B, Immunohistological double labeling of DRG neurons for Kv9.1 and Kv2.1 (left) or Kv2.2 (right) proteins. Examples of colocalization in large neurons are denoted by arrows. Scale bars: 50 μm.

Figure 3.

Kv9.1 is rapidly and robustly downregulated in DRG after nerve injury. A, Animals subjected to sham surgery retain Kv9.1 expression (i), while Kv9.1 mRNA signal is virtually absent in L5 DRG at 14 d postaxotomy (ii). All axotomized L5 DRG neurons lacking Kv9.1 signal are stained positive for the nerve injury marker ATF3 (iii). In the adjacent L4 only a minority of neurons feature loss of Kv9.1 and these are also immunoreactive for ATF3 (arrows in iv). B, C, Images and quantification of Kv9.1 transcriptional downregulation after SNT, illustrating a rapid attenuation of mRNA levels in axotomized L5 DRG as soon as 1 d postinjury, with minimal expression detected as long as day 28 (mean ± SEM; n = 4 per time point; ***p < 0.001 vs sham, one-way ANOVA with Tukey's post hoc). D, Kv9.1 protein staining is reduced 7 d postinjury in lumbar DRG neurons. E, F, Quantification of Kv9.1 protein positivity and immunoreactivity (IR), respectively, showing Kv9.1 protein downregulation in medium-large neurons (mean ± SEM; n = 3–4 per time point; **p < 0.01, ***p < 0.001 vs sham, one-way ANOVA with Tukey's post hoc). Scale bars: 50 μm.

Figure 4.

Kv9.1 is also downregulated in DRG culture. A, Kv9.1 ISH combined with IHC for NF200 and CGRP in dissociated DRG culture demonstrates Kv9.1 mRNA reduction in large neurons (arrows) after 72 h in culture, compared with baseline (3 h postplating). B, Quantification of Kv9.1 mRNA in vitro expression by qRT-PCR (mean ± SEM; n = 3 per dataset, **p < 0.01, t test).

Figure 5.

Intrathecal Kv9.1 siRNA treatment induces pain behaviors in naive rats. A, qRT-PCR quantification of Kv9.1 mRNA in rat PASMC cultures transfected with one of three Kv9.1 siRNA sequences or control siRNA (control, n = 6; siRNA, n = 3 per group; *p < 0.05 vs control, one-way ANOVA with Tukey's). B, qRT-PCR showing Kv9.1 in vivo knock-down in L5 DRG, 4 d after intrathecal delivery of siRNA #1 compared with vehicle or matched scrambled control (vehicle, n = 4; scrambled, n = 5; Kv9.1, n = 7; *p < 0.05, t test). C, IHC for Kv9.1 in scrambled- and siRNA-treated DRG to determine protein knockdown. Graphs illustrate quantification of number of positive myelinated neurons and mean Kv9.1 signal intensity (scrambled, n = 4; siRNA, n = 6; **p < 0.01, ***p < 0.001, t test). D, Kv9.1 siRNA infusion inflicts a reduction in mechanical pain withdrawal thresholds (Kv9.1, n = 7; control, n = 6; *p < 0.05, **p < 0.01, ***p < 0.001 vs scrambled control or baseline, two-way repeated measurements ANOVA with Tukey's). E, There was no change in heat pain thresholds after siRNA treatment. Vertical arrows on x-axis denote siRNA injections. All data represent mean ± SEM.

Figure 6.

Kv9.1 knock-down triggers ectopic activity and a form of peripheral wind-up in response to stimulation. A, Schematic illustrating the positions of stimulating and recording electrodes. B, Example recordings from centrally disconnected L4/L5 strands demonstrating SA in Kv9.1 siRNA-treated or nerve-injured rats, but not in control (scrambled siRNA) animals. C, Frequency-dependent SEA (denoted by double arrowheads) in Kv9.1 siRNA-treated (middle) and injured (right), but not control (left) animals. This activity is not locked in time and can be seen in between stimulation events (vertical arrows on top of 5 Hz stimulation traces, only first 5 shown). Also note the prolonged after-discharge (AD) observed in siRNA-treated and injured animals. D, Percentage of units showing SA and SEA in control (n = 269), Kv9.1 siRNA-treated (n = 369) and injured (n = 176) animals (*p < 0.05, **p < 0.01, ***p < 0.001 vs control, χ2 test). E, Firing rate of SEA units at different stimulation frequencies (mean ± SEM; control, n = 4; siRNA, n = 22; injured, n = 17; *p < 0.05 vs control, two-way ANOVA with Tukey's). F, Quantification of AD rate per SEA unit (mean ± SEM; *p < 0.05 vs control, Mann–Whitney test).

Figure 7.

SEA in siRNA-treated animals is exaggerated by C-fiber activation. Recording from an siRNA-treated animal strand, featuring a small unit firing spontaneously (arrowhead). After stimulation at A-fiber strength (2 Hz), a second previously silent unit is recruited and shows progressive SEA (top, double arrowheads). This phenomenon was exaggerated when delivering stimuli capable of activating C-fibers, even at low frequencies (1 Hz, bottom).

Figure 8.

Intracellular APs from naive, injured (3–5 d postaxotomy), scrambled siRNA-treated, and Kv9.1 siRNA-treated neurons and the effect of elevated [K+]e on excitability. A, Left, Representative APs evoked by dorsal root stimulation from a naive (gray) or an injured (black) cell. Arrow indicates the stimulus artifact. Right, Responses to a series of depolarizing current injections from a naive and an injured neuron at 3, 6, and 9 mm [K+]e. B, Left, Representative APs from scrambled- and Kv9.1 siRNA-treated neurons. The inset highlights the shortened AHP in Kv9.1 siRNA-treated neuron on a different scale. Right, neuronal responses to a series of depolarizing current injections from control or Kv9.1 siRNA-treated neurons at 3, 6, and 9 mm [K+]e. C, Comparison of frequency distribution of threshold current at elevated [K+]e (6–9 mm) between naive/injured neurons (left) and scrambled/Kv9.1 siRNA-treated neurons (right). Dashed lines indicate the respective cumulative percentages, illustrating the shift toward lower values in injured and Kv9.1 siRNA-treated neurons. D, representative AP from a naive neuron in the absence (black) or presence (gray) of the Kv2 blocker stromatoxin-1 (ScTx). The inset highlights the shortened AHP caused by ScTx, similar to Kv9.1 siRNA treatment.