The MNK-eIF4E Signaling Axis Contributes to Injury-Induced Nociceptive Plasticity and the Development of Chronic Pain

Abstract

Injury-induced sensitization of nociceptors contributes to pain states and the development of chronic pain. Inhibiting activity-dependent mRNA translation through mechanistic target of rapamycin and mitogen-activated protein kinase (MAPK) pathways blocks the development of nociceptor sensitization. These pathways convergently signal to the eukaryotic translation initiation factor (eIF) 4F complex to regulate the sensitization of nociceptors, but the details of this process are ill defined. Here we investigated the hypothesis that phosphorylation of the 5' cap-binding protein eIF4E by its specific kinase MAPK interacting kinases (MNKs) 1/2 is a key factor in nociceptor sensitization and the development of chronic pain. Phosphorylation of ser209 on eIF4E regulates the translation of a subset of mRNAs. We show that pronociceptive and inflammatory factors, such as nerve growth factor (NGF), interleukin-6 (IL-6), and carrageenan, produce decreased mechanical and thermal hypersensitivity, decreased affective pain behaviors, and strongly reduced hyperalgesic priming in mice lacking eIF4E phosphorylation (eIF4E^S209A ). Tests were done in both sexes, and no sex differences were found. Moreover, in patch-clamp electrophysiology and Ca²⁺ imaging experiments on dorsal root ganglion neurons, NGF- and IL-6-induced increases in excitability were attenuated in neurons from eIF4E^S209A mice. These effects were recapitulated in Mnk1/2^-/- mice and with the MNK1/2 inhibitor cercosporamide. We also find that cold hypersensitivity induced by peripheral nerve injury is reduced in eIF4E^S209A and Mnk1/2^-/- mice and following cercosporamide treatment. Our findings demonstrate that the MNK1/2-eIF4E signaling axis is an important contributing factor to mechanisms of nociceptor plasticity and the development of chronic pain.SIGNIFICANCE STATEMENT Chronic pain is a debilitating disease affecting approximately one in three Americans. Chronic pain is thought to be driven by changes in the excitability of peripheral nociceptive neurons, but the precise mechanisms controlling these changes are not elucidated. Emerging evidence demonstrates that mRNA translation regulation pathways are key factors in changes in nociceptor excitability. Our work demonstrates that a single phosphorylation site on the 5' cap-binding protein eIF4E is a critical mechanism for changes in nociceptor excitability that drive the development of chronic pain. We reveal a new mechanistic target for the development of a chronic pain state and propose that targeting the upstream kinase, MAPK interacting kinase 1/2, could be used as a therapeutic approach for chronic pain.

Keywords: MNK1; MNK2; chronic pain; dorsal root ganglion; eIF4E; nociceptor.

Figure 1.

eIF4E^S209A mice have normal acute nociceptive responses but decreased mechanical hypersensitivity to formalin and a group I mGluR agonist. A, B, eIF4E^S209A and WT mice show no differences in tail-flick responses (55°C; n ≥ 6; A) or baseline (BL) paw withdrawal thresholds (n ≥ 6; B). C, D, First phase (0–10 min) and second phase (15–45 min) summed responses were not different between eIF4E^S209A and WT mice (n ≥ 10). E, Three days after intraplantar injection of 5% formalin, eIF4E^S209A mice exhibited a significantly higher mechanical withdrawal threshold compared with WT mice (n ≥ 10). F, G, The mGlu1/5 receptor agonist DHPG (50 nmol) was injected intrathecally in both WT and eIF4E^S209A mice. Nocifensive behaviors summed during the 30 min after injection were equal in both strains (n = 8). However, 6 h after DHPG intrathecal injection, WT mice exhibited mechanical hypersensitivity, whereas eIF4E^S209A mice did not (n = 8). **p < 0.01; ****p < 0.0001.

Figure 2.

Normal development of nociceptive pathways in eIF4E^S209A mouse DRGs and spinal cord. A, B, WT and eIF4E^S209A mouse spinal cords were immunostained with CGRP (cyan), TRPV1 (green), and IB₄ (red; representative images from n = 3 mice). C, Immunostaining for peripherin (green) and NF200 (red) in DRGs from WT and eIF4E^S209A mice shows no differences in the proportion of peripherin-positive neurons per section (n = 3), the proportion of NF200-positive neurons per section (n = 3) or in overlap between the two markers (n = 3). D, WT and eIF4E^S209A DRGs were stained for TRPV1 (red), IB₄ (green), and CGRP (blue), and the proportion of neurons expressing each marker was assessed (TRPV1: WT mean = 34.3 ± 3.1% vs eIF4E^S209A mean = 32.9 ± 1.3%; IB₄: WT mean = 19.6 ± 1.8% vs eIF4E^S209A mean = 17.8 ± 2.0%; WT mean = CGRP: 29.1 ± 1.5% vs eIF4E^S209A mean = 31.5 ± 2.0%). TRPV1 and IB₄ populations were segregated in both mouse strains (TRPV1/IB₄ overlap: WT mean = 3.7 ± 0.1% vs eIF4E^S209A mean = 3.2 ± 0.7%), while TRPV1 and CGRP overlapped substantially (TRPV1/CGRP overlap: WT mean = 74.2 ± 6.3% vs eIF4E^S209A mean = 71.1 ± 2.3%) as demonstrated in the proportional Venn diagrams. Scale bars, 200 μm.

Figure 3.

Normal ERK and 4E-BP phosphorylation in eIF4E^S209A mouse DRGs and spinal cord. A, B, eIF4E^S209A DRG shows equal levels of ERK and 4E-BP phosphorylation, while eIF4E phosphorylation is completely absent compared with WT DRG using Western blot analysis (n ≥ 6). C, D, Additionally, spinal cord from eIF4E^S209A shows similar levels of p-ERK and p-4E-BP compared with WT spinal cord (n ≥ 5).

Figure 4.

Mechanical and thermal hypersensitivity, facial grimacing, and the development of hyperalgesic priming are decreased in eIF4E^S209A mice. A, IL-6 (0.1 ng) was injected into the hindpaw in both WT and eIF4E^S209A mice. Hindpaw mechanical thresholds were measured at 3, 24, 48, and 72 h. B, eIF4E^S209A mice exhibited reduced mechanical hypersensitivity compared with WT mice, and a deficit in hyperalgesic priming (n ≥ 6). C–F, eIF4E^S209A mice also demonstrated decreased mechanical (C) and thermal (D; n = 6) hypersensitivity in response to intraplantar injection of 50 ng of NGF (n ≥ 6) and NGF-induced hyperalgesic priming (E, F). G, Intraperitoneal injection of 20 ng of 2at-LIGRL likewise induced decreased mechanical hypersensitivity in eIF4E^S209A mice and a decrease of hyperalgesic priming in response to IL-6 (H; n ≥ 6). I, J, 2at-LIGRL induces facial grimacing in WT mice but not in eIF4E^S209A mice (I); moreover, eIF4E^S209A mice fail to show facial grimacing after 2at-LIGRL priming when subsequently challenged with PGE₂ (n ≥ 6; J). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. BL, Baseline.

Figure 5.

MNK1/2–eIF4E signaling contributes to the development of mechanical and thermal hypersensitivity and hyperalgesic priming in response to inflammatory stimuli. A, B, Carrageenan (0.5% w/v) was injected into the hindpaw in both WT and eIF4E^S209A mice. Hindpaw mechanical and thermal thresholds show that eIF4E^S209A mice exhibited a blunted mechanical and thermal hypersensitivity compared with WT mice (n ≥ 5). C, eIF4E^S209A mice developed reduced hyperalgesic priming when injected with PGE₂ (n ≥ 5). D, E, Mnk1/2^−/− mice injected with CFA (0.5 mg/ml, i.pl.; 10 μl) recover faster in both mechanical and thermal hypersensitivity compared with WT mice (n = 9). F, G, Mnk1/2^−/− mice show decreased mechanical and thermal response to PGE₂ after recovering from the initial hypersensitivity from CFA (n = 9). *p < 0.05; **p < 0.01; ***p < 0.001. BL, Baseline.

Figure 6.

Local inhibition of MNK1/2 with cercosporamide reduces eIF4E phosphorylation, mechanical hypersensitivity, grimacing, and inhibits the development of hyperalgesic priming. A–C, In vitro treatment for 1 h of DRG neurons with cercosporamide (10 μm) decreased eIF4E phosphorylation (n = 3; A) but did not influence 4E-BP1 phosphorylation (B; n = 3) or ERK phosphorylation (C; n = 3). D, E, Injection of cercosporamide (40 mg/kg, i.p.) results in a decreased eIF4E phosphorylation in DRGs (n = 3; D) but does not affect 4E-BP1 phosphorylation (E; n = 3). F, G, Mechanical hypersensitivity induced by NGF (50 ng) or 2at-LIGRL (20 ng) was reduced by cercosporamide (10 μg, n ≥ 9). Similar to eIF4E^S209A mice, pharmacological inhibition of MNK1/2 using cercosporamide attenuated the development of hyperalgesic priming induced by NGF (H) and 2at-LIGRL (I; n ≥ 5). J, K, Facial grimacing induced by the injection of 2at-LIGRL (20 ng, i.pl.) was also attenuated with local cercosporamide (10 μg) treatment (J) as was facial grimacing with subsequent PGE₂ challenge (K; n ≥ 5). L, Immunostaining of glabrous skin for TRPV1 (red) and p-eIF4E (green) revealed that eIF4E phosphorylation is present in TRPV1-positive fibers in WT mice, but is absent in eIF4E^S209A mice. Scale bar, 20 μm. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. BL, Baseline; Cerco, cercosporamide.

Figure 7.

MNK1/2–eIF4E signaling mediates NGF- and IL-6-induced changes in excitability in DRG neurons. A, WT DRG neurons were exposed to NGF or vehicle 18–24 h before patch-clamp recordings. Ramp current-evoked spiking demonstrates that NGF exposure increases the excitability of WT DRG neurons. B, eIF4E^S209A DRG neurons showed no difference in the number of spikes evoked by ramp currents between NGF and vehicle-treated DRG neurons. C, Membrane capacitance measures between WT and eIF4E^S209A DRG neurons show no difference in neuron size between samples demonstrating that small-diameter neurons were used for recordings. D, Pharmacological inhibition of MNK1/2 using cercosporamide (10 μm, n ≥ 9) recapitulated the effect seen in eIF4E^S209A DRG neurons blocking NGF-induced hyperexcitability. E, F, IL-6 (50 ng/ml) exposure to WT DRG neurons for 1 h caused an increase in excitability compared with vehicle (E) but failed to do so in eIF4E^S209A DRG neurons (F). G, Membrane capacitance between these samples was not different. N for each condition is shown in the appropriate bar. H, Cercosporamide blocked enhanced excitability induced by IL-6 in small-diameter DRG neurons (50 ng/ml; n ≥ 7). Traces shown in all panels are for the 700 pA stimulus. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Cerco, Cercosporamide; VEH, vehicle.

Figure 8.

Cercosporamide has no additional effect in eIF4E^S209A mice and no acute effect on sodium current density. A, Mechanical hypersensitivity induced by NGF (50 ng) showed no difference between eIF4E^S209A mice that additionally received a hindpaw injection of cercosporamide (10 μg) and eIF4E^S209A mice that did not (n ≥ 4; p > 0.05, two-way ANOVA). B, Subsequent injection of PGE₂ to precipitate priming also shows no difference between eIF4E^S209A mice that previously received cercosporamide and eIF4E^S209A mice that did not (n ≥ 4; p > 0.05, two-way ANOVA). C, D, Sodium current (C) and sodium current density (D) were measured in WT DRG neurons with cercosporamide (10 μm) perfused acutely over a 5–10 s period (n = 4; p > 0.05, two-way ANOVA), indicating the cercosporamide does not block voltage-gated sodium channels at this concentration. BL, Baseline; Cerco, cercosporamide; VEH, vehicle.

Figure 9.

IL6-induced sustained spiking and increased voltage-gated sodium current density is altered in eIF4E^S209A DRG neurons. A, Image of DRG neurons cultured on an MEA device. Scale bar, 200 μm. B, Raster plot showing electrical activity of WT and eIF4E^S209A DRG neuronal networks observed using MEAs during baseline, IL-6 treatment, and 1 h wash. C, IL-6 elicited increased spiking in WT DRG cultures during treatment that was sustained throughout the washout period and was significantly greater than the spiking observed in eIF4E^S209A DRG neurons. D, E, Sodium current density was increased after 1 h of IL-6 exposure in WT DRG neurons but was decreased in eIF4E^S209A DRG neurons (WT, n ≥ 7; eIF4E^S209A, n ≥ 12). *p < 0.05, ***p < 0.001.

Figure 10.

IL-6 treatment alters PGE₂ responsiveness in DRG neurons in an eIF4E phosphorylation-dependent fashion. A, B, WT and eIF4E^S209A DRG neurons were cultured and treated with vehicle or IL-6 (50 ng/ml) for 1 h. PGE₂ (1 nm) was perfused for 30 s, during which Ca²⁺ responses were measured. C, IL-6 treatment decreased the latency of PGE₂-evoked Ca²⁺ release in WT DRG neurons compared with vehicle-treated WT neurons (n ≥ 50). D, The proportion of neurons responding to PGE₂ was increased after IL-6 exposure in WT DRG neurons but was unchanged in eIF4E^S209A DRG neurons. Scale bar, 20 μm. ****p < 0.0001. VEH, Vehicle.

Figure 11.

Normal Ca²⁺ signaling evoked by KCl and capsaicin in eIF4E^S209A DRG neurons. A, WT and eIF4E^S209A DRG neurons were imaged after 1 h of treatment with vehicle or IL-6 (50 ng) during vehicle-containing bath solution and capsaicin (250 nm)-containing bath solution perfusions. B, The proportion of TRPV1-positive neurons was unchanged in WT and eIF4E^S209A DRG neurons treated with vehicle of IL-6. C, Moreover, Ca²⁺ signaling evoked by KCl (50 mm) was unchanged in all conditions (n ≥ 57 neurons/condition; p > 0.05, one-way ANOVA). Scale bar, 20 μm. VEH, Vehicle.

Figure 12.

Decreased neuropathic pain in eIF4E^S209A mice and through MNK1/2 inhibition. A, Following SNI surgery eIF4E^S209A mice show reduced mechanical hypersensitivity that normalizes 14 d following surgery compared with WT mice (n = 8). B, Mnk1/2^−/− mice show blunted mechanical hypersensitivity lasting 35 d after SNI surgery (n = 10). C, Following SNI eIF4E^S209A mice have a decrease in cold hypersensitivity as measured in the acetone test compared with WT mice (n = 8). D, Mnk1/2^−/− mice show a sustained decrease in cold hypersensitivity after SNI compared with WT mice (n = 10). E, Following systemic administration of cercosporamide (40 mg/kg) for 3 d, WT mice show a transient decrease in cold hypersensitivity (n = 4). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. BL, Baseline; Cerco, cercosporamide.