Nociceptor-specific gene deletion reveals a major role for Nav1.7 (PN1) in acute and inflammatory pain

Abstract

Nine voltage-gated sodium channels are expressed in complex patterns in mammalian nerve and muscle. Three channels, Na(v)1.7, Na(v)1.8, and Na(v)1.9, are expressed selectively in peripheral damage-sensing neurons. Because there are no selective blockers of these channels, we used gene ablation in mice to examine the function of Na(v)1.7 (PN1) in pain pathways. A global Na(v)1.7-null mutant was found to die shortly after birth. We therefore used the Cre-loxP system to generate nociceptor-specific knockouts. Na(v)1.8 is only expressed in peripheral, mainly nociceptive, sensory neurons. We knocked Cre recombinase into the Na(v)1.8 locus to generate heterozygous mice expressing Cre recombinase in Na(v)1.8-positive sensory neurons. Crossing these animals with mice where Na(v)1.7 exons 14 and 15 were flanked by loxP sites produced nociceptor-specific knockout mice that were viable and apparently normal. These animals showed increased mechanical and thermal pain thresholds. Remarkably, all inflammatory pain responses evoked by a range of stimuli, such as formalin, carrageenan, complete Freund's adjuvant, or nerve growth factor, were reduced or abolished. A congenital pain syndrome in humans recently has been mapped to the Na(v)1.7 gene, SCN9A. Dominant Na(v)1.7 mutations lead to edema, redness, warmth, and bilateral pain in human erythermalgia patients, confirming an important role for Na(v)1.7 in inflammatory pain. Nociceptor-specific gene ablation should prove useful in understanding the role of other broadly expressed genes in pain pathways.

Fig. 1.

Generation of Na_v1.8Cre mice. (A) Diagram of the native Na_v1.8 allele, Na_v1.8Cre targeting construct, and Na_v1.8Cre-targeted allele before and after excision of neo^r. (B and C) Histological localization of Na_v1.8Cre-mediated β-galactosidase activity. Serial 10-μm sections from 4% paraformaldehydefixed Na_v1.8Cre^+/-,R26R^+/- animals were stained with antibodies to β-galactosidase (B) and with anti-Na_v1.8 polyclonal antisera (C). Arrows highlight double-labeled cells. (Scale bars, 50 μm.) (D) Southern blot with BamHI and external probe confirms correct targeting and excision of neo^r. The 6.5-kb WT band is seen in all lanes, and the 8.5- and 7.3-kb bands represent the targeted allele before (+neo^r) and after (-neo^r) excision of neo^r, respectively.

Fig. 2.

Generation and analysis of Na_v1.7 floxed mice. (A) Structure of the native Na_v1.7 allele, Na_v1.7 targeting construct, fNa_v1.7 allele, fNa_v1.7 allele after excision of neo^r, and Na_v1.7 knockout allele. (B) Southern blotting with EcoRI and the external probe confirms correct targeting. (C) Southern blotting with ApaI and internal probe confirms the removal of neo^r cassette. (D) Southern blotting confirms the deletion of exons 14 and 15 in Na_v1.7^-/-. (E) PCR was used to detect exon 14-15 deletion in genomic and cDNA from DRG but not spinal cord in Na_v1.7R^-/-.

Fig. 3.

Electrophysiology of Na_v1.7^-/-.(A) TTX-sensitive and -resistant peak sodium current density in Na_v1.7R^-/- (black) (n = 7) and fNa_v1.7 (white) (n = 14) mice. (B) TTX-resistant currents in WT (hatched) (n = 22), Na_v1.8Cre heterozygotes (filled) (n = 22), and homozygous (n = 26) mice. (C-F) Responses of lamina V dorsal horn neurons (L3 and L4) receiving input from the hindpaw to peripheral stimuli in Na_v1.7 R^-/- (black) (n = 14) and littermate fNa_v1.7 (white) (n = 18) mice. (C) Evoked responses to von Frey hairs showed a mechanical deficit in Na_v1.7R^-/- mice (P = 0.048). Temperature (water jet) (D) and pinch, brush, and noxious cold (E) applied to the peripheral receptive field over 10 s were normal. Data are expressed as mean spikes (±SEM) evoked over 10 s. (F) Response to transcutaneous electrical stimulation of the receptive field (train of 16, 2-ms-wide electrical pulses, at 0.5 Hz, at three times C-fiber threshold). Spikes evoked between 0 and 50 ms were classified as caused by A-fiber input, and those evoked between 50 and 250 ms were classified as caused by C-fiber input. “Input” is the number of spikes evoked by the first stimulus of the train (50-800 ms) and reflects nonpotentiated C-fiber-evoked dorsal horn neuron response. Excess spikes, additional spikes recorded above the predicted constant baseline response, are a measure of “wind-up,” defined as increased neuronal excitability to repeated constant stimulation.

Fig. 4.

Acute pain behavior of Na_v1.7R^-/- mice. (A) Noxious thermal stimulation of fNa_v1.7 (white) and Na_v1.7R^-/- (black) mice by using Hargreave's apparatus. The latency of hindpaw withdrawal was significantly increased in Na_v1.7R^-/- mice (11.01 ± 0.47, n = 14, and 8.23 ± 0.67, n = 12, respectively) (P = 0.003, t test). (B) Response to noxious thermal stimulation by using the hotplate apparatus was not significantly different in fNa_v1.7 (n = 12) and Na_v1.7R^-/- (n = 17) mice (P = 0.27, 0.11, and 0.79, t test). (C)Na_v1.7R^-/- mice (396.4 ± 4.0, n = 19) showed profound analgesia to noxious mechanical pressure when using the Randall-Selitto apparatus compared with fNa_v1.7 mice (195.5 ± 4.8, n = 18) (P < 0.0001, t test). (D) Response to mechanical stimulation when using von Frey hairs was not significantly different in fNa_v1.7 (n = 11) and Na_v1.7R^-/- (n = 19) mice.

Fig. 5.

Inflammatory pain in Na_v1.7R^-/-. (A) Analysis of behavior of fNa_v1.7 (white) and Na_v1.7R^-/- (black) mice in models of inflammatory pain. (Aa) Pain behavior of fNa_v1.7 (n = 14) and Na_v1.7R^-/- (n = 17) mice after intraplantar injection of 20 μl of 5% formalin. Time spent licking/biting the injected hindpaw in phase I (1-10 min) and phase II (10-55 min) was recorded. There is a reduction in phase I (P = 0.03, t test) (fNa_v1.7, 93.1 ± 7.2; Na_v1.7R^-/-, 70.4 ± 6.6). In phase II, Na_v1.7R^-/- mice also showed a significant reduction in pain behavior (P = 0.004, t test) (101.3 ± 18.9 and 220.9 ± 34.8 respectively). (Ab) Time course of the formalin response (P < 0.05 at 5-, 20-, 25-, and 30-min time points). (Ac) Thermal hyperalgesia after intraplantar injection of 20 μl of CFA. fNa_v1.7 mice developed pronounced hyperalgesia (0.56 ± 0.1, n = 9) from day 1, whereas Na_v1.7R^-/- mice did not (0.97 ± 0.06, n = 14) (P < 0.001, ANOVA; P < 0.05 at all time points). (Ad) Mechanical allodynia induced by intraplantar injection of 20 μl of CFA. fNa_v1.7 mice developed pronounced allodynia from day 1 (0.50 ± 0.16, n = 8), whereas the Na_v1.7R^-/- mice did not (1.05 ± 0.12, n = 13) (P = 0.012, ANOVA; P < 0.05 at days 2, 3, 4, and 10). (Ae) Thermal hyperalgesia after intraplantar injection of 20 μl of 2% carrageenan. fNa_v1.7 mice developed pronounced hyperalgesia (0.47 ± 0.03, n = 7) within an hour, whereas Na_v1.7R^-/- mice did not (1.07 ± 0.04, n = 5) (P < 0.001, ANOVA; P < 0.05 at all time points). (Af) Thermal hyperalgesia after intraplantar injection of 500 ng of NGF. fNa_v1.7 mice (n = 15) developed thermal hyperalgesia in a biphasic pattern, whereas Na_v1.7R^-/- mice (n = 8) showed no phase I (first hour) and a reduced phase II (P < 0.05 at 0.25, 0.5, 1, 2, 6.5, and 8 h, ANOVA). (B) Inflammatory pain is normal in Na_v1.8Cre mice. Analysis of behavior of Na_v1.8Cre^+/- (black) and C57BL/6 (white) mice in models of inflammatory pain. (Ba) Behavior of Na_v1.8Cre^+/- (n = 6) and C57BL/6 (n = 6) mice after intraplantar injection of 20 μl of 5% formalin was not different between groups in phase I (107 ± 12.6 s and 96.1 ± 10.3 s, respectively; P = 0.51) and phase II (230 ± 47.7 and 290 ± 45.4 s, respectively; P = 0.38, t test). (Bb) Thermal hyperalgesia after intraplantar injection of 20 μl of CFA was not different in Na_v1.8Cre^+/- (0.58 ± 0.05, n = 6) and WT (0.55 ± 0.05, n = 6) littermates (P = 0.79, ANOVA). (Bc) Mechanical allodynia after intraplantar injection of 20 μl of CFA was not different in Na_v1.8Cre^+/- (0.30 ± 0.16, n = 6) and WT (0.27 ± 0.04, n = 6) littermates (P = 0.74, ANOVA). (Bd) Thermal hyperalgesia induced by intraplantar injection of 20 μl of 2% carrageenan was not different in Na_v1.8Cre^+/- (0.40 ± 0.26, n = 9) and C57BL/6 (0.36 ± 0.03, n = 7) mice (P = 0.31, two-way ANOVA). (Be) Thermal hyperalgesia induced by intraplantar injection of 500 ng of NGF was not different in Na_v1.8Cre^+/- (0.58 ± 0.02) (n = 6) and C57BL/6 (0.56 ± 0.03) (n = 6) mice (P = 0.83, two-way ANOVA).