Functional Coupling of Slack Channels and P2X3 Receptors Contributes to Neuropathic Pain Processing

Abstract

The sodium-activated potassium channel Slack (K_Na1.1, Slo2.2, or Kcnt1) is highly expressed in populations of sensory neurons, where it mediates the sodium-activated potassium current (I_KNa) and modulates neuronal activity. Previous studies suggest that Slack is involved in the processing of neuropathic pain. However, mechanisms underlying the regulation of Slack activity in this context are poorly understood. Using whole-cell patch-clamp recordings we found that Slack-mediated I_KNa in sensory neurons of mice is reduced after peripheral nerve injury, thereby contributing to neuropathic pain hypersensitivity. Interestingly, Slack is closely associated with ATP-sensitive P2X3 receptors in a population of sensory neurons. In vitro experiments revealed that Slack-mediated I_KNa may be bidirectionally modulated in response to P2X3 activation. Moreover, mice lacking Slack show altered nocifensive responses to P2X3 stimulation. Our study identifies P2X3/Slack signaling as a mechanism contributing to hypersensitivity after peripheral nerve injury and proposes a potential novel strategy for treatment of neuropathic pain.

Keywords: P2X3; Slack; dorsal root ganglia; mice; neuropathic pain.

Figure 1

Neuropathic pain behavior is increased in Slack^-/- mice. (A) paw withdrawal latencies of Slack^-/- and wild-type (WT) mice after mechanical stimulation with von Frey filaments (up-and-down method) in the spared nerve injury (SNI) model of neuropathic pain (n = 10 animals per group). Thirteen days after SNI, Slack^-/- mice showed increased mechanical hypersensitivity compared to WT littermates (two-way ANOVA; p = 0.0312; WT versus Slack^-/-). (B) percentage of weight bearing on the ipsilateral hind paw relative to both hind paws in Slack^-/- and WT mice (n = 10 animals per group), as assessed using a dynamic weight-bearing device. Twelve days after SNI, the weight-bearing reduction was more pronounced in Slack^-/- mice compared to WT littermates (two-way ANOVA; p = 0.0464; WT versus Slack^-/-) Bars denote mean ± SEM. * p ˂ 0.05.

Figure 2

Slack-mediated potassium currents in sensory neurons are reduced after SNI. (A,B) Representative outward K⁺ current (I_K) traces (A) and associated current-voltage (I-V) curves (B) from whole-cell voltage recordings on IB4-positive lumbar (L4-L5) dorsal root ganglion (DRG) neurons of WT (black) and Slack^-/- mice (red) 14–19 days after spared nerve injury (SNI). Contralateral DRG neurons were used as control. Recordings shown in (A) and (B) were performed in the presence of 140 mM NaCl in the external solution, i.e., under physiological conditions. n = 21–29 cells per group. Repeated ANOVA measures followed by Fisher’s Least Significant Difference test; WT control versus WT SNI: p = 0.0062; Slack^-/- control versus Slack^-/- SNI: p = 0.4374. (C,D) Representative I_K traces (C) and associated I-V curves (D) in the same experimental setting as shown in (A) and (B), however, after replacement of NaCl by 140 mM choline chloride in the external solution to obtain Na⁺-free conditions. n = 7–9 cells per group. Repeated ANOVA measures: WT control versus WT SNI: p = 0.1825; Slack^-/- control versus Slack^-/- SNI: p = 0.6125. The data show that Na⁺-activated I_K (I_KNa) in sensory neurons is carried by Slack channels and reduced after SNI. Data in (B) and (D) are mean ± SEM. * p ˂ 0.05.

Figure 3

Unaltered Slack expression in WT mice after peripheral nerve injury. (A,B) Double-labeling immunostaining of Slack and the neuronal marker anti-βIII-tubulin (TUBB3) in DRGs of naive WT mice and 14 days after SNI. Colocalization of Slack and TUBB3 appears in yellow. The staining suggests that Slack is exclusively localized to neurons and that its distribution is not altered in response to the injury. The absence of Slack immunoreactivity in DRGs of Slack^-/- mice confirms the antibody specificity (A). The percentage of Slack-immunoreactive DRG neurons of all βIII-tubulin-stained neurons is similar in naive mice and 14 days after SNI ((B); 1833 and 1630 cells counted, respectively; n = 3 mice per group). Student’s t-test: p = 0.9130. (C), Quantitative RT-PCR experiments showed that Slack mRNA levels are not altered in DRGs 7 or 14 days after SNI as compared to naive control animals (n = 8 mice per group). One-way ANOVA: p = 0.1433. (D), Immunostaining in the lumbar spinal cord of WT mice 14 d after SNI shows similar Slack expression in the ipsilateral and contralateral dorsal horn. The absence of Slack immunoreactivity in the spinal cord of Slack^-/- mice confirms the antibody specificity. (E,F) A Western blot of spinal cord extracts shows similar Slack protein (140 kDa) expression in naive mice and 7 or 14 days after SNI surgery (E). Uncropped original image is shown in Figure S2A. Quantification is shown in (F). Alpha-tubulin was used as a loading control. One-way ANOVA: p = 0.4446. Bars denote mean ± SEM. Scale bars: 50 µm (A), 200 µm (D).

Figure 4

Slack channels co-localize with P2X3 receptors in sensory neurons. (A–C) Double-labeling immunostaining of Slack and P2X3 in sensory neurons (A) revealed that 97.1% ± 0.2% of Slack-positive DRG neurons co-stained with P2X3 ((B); 1203 cells counted, n = 3 mice) and that 94.6% ± 2.0% of P2X3-positive DRG neurons co-stained with Slack ((C); 1203 cells counted, n = 3 mice). (D) Double-labeling immunostaining of Slack and P2X3 in the spinal cord indicates a high degree of co-localization in the superficial dorsal horn. (E,F) Western blot of P2X3 in spinal cord (SC) and DRGs from WT and Slack^-/- mice demonstrates identical abundance of P2X3 in both genotypes. The uncropped original image is shown in Figure S2B. Student’s t-test: p = 0.5986 in the spinal cord and p = 0.7631 in DRGs. Alpha-tubulin was used as a loading control. (G) Immunostaining revealed that the percentage of DRG neurons positive for P2X3 is similar in WT and Slack^-/- mice. Student’s t-test: p = 0.4046. Bars denote mean ± SEM. Scale bars: 50 µm (A) and 100 µm (D).

Figure 5

Slack-mediated potassium currents are altered by P2X3 activation in vitro. Representative I_K traces from whole-cell voltage-clamp recordings on HEK-Slack and HEK-Slack-P2X3 cells are shown. The current traces presented in (A–C) were recorded in the presence of 2 mM CaCl₂ in the external solution (n = 9–10 cells per group), whereas those depicted in (D–F) were recorded after the replacement of CaCl₂ by MgCl₂ (n = 5–8 cells per group). Experiments were performed without (control) or with the addition of the P2X3 agonist α,β-methylene ATP (α,β-meATP; 30 µM) to the external solution. Note that the P2X3 agonist exerted opposite effects in HEK-Slack-P2X3 cells dependent on the Ca²⁺ concentration: α,β-meATP reduced I_K in the presence of Ca²⁺, whereas it increased I_K under Ca²⁺-free conditions. Data are shown as mean ± SEM. Student’s t-test, * p < 0.05.

Figure 6

P2X3-mediated Ca²⁺ traces are normal in sensory neurons of Slack^-/- mice. (A) Representative examples of Fura-2 ratiometric Ca²⁺ traces in DRG neurons of WT and Slack^-/- mice evoked by α,β-meATP (30 µM, 30 s application) and KCl (75 mM, 20 s application). Responses to KCl were used to test neuron viability. Experiments were performed in lumbar (L4-L5) DRG neurons (n = 402–518 neurons per group). (B,C) Quantification of the magnitude of the Ca²⁺ response to α,β-meATP stimulation with percentage above baseline ((B) p = 0.7858) and ratio peak ((C) p = 0.7858). (D) Quantification of the percentage of responsive neurons to α,β-meATP stimulation (p = 0.5966). The data show that α,β-meATP-evoked Ca²⁺ responses are similar in DRG neurons from Slack^-/- and WT mice. Bars denote mean ± SEM and circles show data from each neuron in (B) and from each mouse in (D). Student’s t-tests were performed.

Figure 7

P2X3-dependent nocifensive behavior is altered in Slack^-/- mice. Paw-licking responses induced by paw injection of drugs are shown. (A) The immediate paw licking in the first 2 min after injection of α,β-meATP (12 nmol) is increased in naive Slack^-/- mice compared with WT mice (n = 7/genotype; p = 0.0398 in 0–2 min; p = 0.9800 in 2–6 min; p = 0.9352 in 6–10 min). (B) No significant differences between groups occurred when the P2X3 receptor antagonist AF353 (70 nmol intraplantar) was injected 10 min prior to α,β-meATP (n = 5–7/genotype; p = 0.9664 in 0–2 min; p = 0.9732 in 2–6 min; p = 0.7546 in 6–10 min). (C) Paw-licking responses after injection of the vehicle were comparable in Slack^-/- and WT mice (n = 6/genotype; p = 0.9999 in 0–2 min; p = 0.8934 in 2–6 min; p = 0.8709 in 6–10 min) and similar to the licking behavior after combined injection of α,β-meATP and AF353 (B). (D) When α,β-meATP (12 nmol) was injected in the ipsilateral hind paw after SNI, the paw-licking response persisted over the 10 min observation period in WT mice. In Slack^-/- mice, the paw licking was increased in the first 2 min and decreased from 6–10 min compared with WT mice (n = 16/genotype; p = 0.0262 in 0–2 min; p = 0.9993 in 2–6 min; and p = 0.0362 in 6–10 min). Two-way ANOVA tests were performed. Bars denote mean ± SEM. * p ˂ 0.05.

Figure 8

Proposed model demonstrating the functional coupling of Slack channels and P2X3 receptors in IB4-positive sensory neurons. Activation of P2X3 receptors by ATP may lead to an influx of Na⁺, Ca²⁺, or both into sensory neurons. (A) P2X3-mediated Na⁺ influx activates Slack, which results in K⁺ efflux and thus partial inhibition of neuronal activity. (B) P2X3-mediated Ca²⁺ influx inhibits Slack, which leads to increased neuronal activity due to the lack of inhibition.