Local translation and retrograde axonal transport of CREB regulates IL-6-induced nociceptive plasticity

The transcription factor CREB is synthesized within axons of developing DRG neurons in an NGF-dependent fashion and transported back to the neuronal nucleus where it influences transcription of genes regulating the survival of DRG neurons [15]. Local translation plays a key role in the sensitization of mammalian nociceptors, and is involved in the initiation and maintenance of hyperalgesic priming. Local translation is likewise linked to hypersensitivity arising in invertebrate sensory neurons after crush injury. We therefore hypothesized that local translation of CREB may link local translation of a transcription factor to IL-6-induced nociceptive plasticity. We have previously demonstrated that IL-6 activates the ERK/eIF4E pathway leading to enhanced translation of proteins in the axons of primary sensory neurons. Moreover, IL-6 induced mechanical hypersensitivity and subsequent PGE₂-induced precipitation of hyperalgesic priming is blocked by inhibition of translation at the site of IL-6 injection [6]. As a first test of our hypothesis we asked if IL-6 induces a rapid change in CREB expression in DRG neurons. Treatment of DRG neurons in culture with IL-6 (50 ng/mL) for 15 minutes caused increased expression of CREB (Figure  1A) consistent with the time and concentration dependent activation of ERK and eIF4E phosphorylation we have determined previously.This finding suggests that DRG neurons can synthesize CREB in response to IL-6. In order to determine if an IL-6-mediated synthesis of CREB in DRG neurons can lead to retrograde transport, we injected IL-6 (0.1 ng) or vehicle into the paws of mice and collected the sciatic nerves at 15 min (a time point too rapid for retrograde transport) or 2 hrs (a time point where maximal retrograde transport would be predicted based on standard retrograde transport rates and the distance from the paw where the sciatic nerve was removed) later. As expected, at 15 min post injection, IL-6 did not induce an increase in the levels of CREB within the sciatic nerve (Figure  1B). However, 2 hrs following i.pl. injection of IL-6 we observed a significant increase in levels of CREB in the sciatic nerve (Figure  1B). These data suggest that locally synthesized CREB in the axons of DRG neurons in the paws of mice that were treated with IL-6 undergo retrograde transport into to the sciatic nerve. We next sought to demonstrate that this IL-6-induced CREB translocation to the sciatic nerve was nascently synthesized at the site of IL-6 injection. To this end we co-injected IL-6 with the methionine analogue AHA along with vehicle or IL-6 and sciatic nerves were dissected after 2 hrs. In order to detect AHA incorporation into nascently synthesized CREB we immunoprecipitated CREB and used “click chemistry” to label AHA with biotin, which was subsequently detected by Western blotting. Using this method, we observed an IL-6-dependent increase in nascently synthesized CREB in the sciatic nerve 2 hrs following treatment (Figure  1C). Collectively, these data provide evidence that IL-6 stimulates the local translation of CREB in axons of DRG neurons, which then undergoes retrograde transport into the sciatic nerve.

Our hypothesis posits that translation links local signaling events to long-term changes in nociceptors via retrograde transport of a locally synthesized transcription factor. Hence, we next sought to determine the role of retrograde axonal transport in the development and maintenance of hyperalgesic priming. To this end, we used colchicine and nocodazole, which disrupt the polymerization of tubulin monomers to form microtubules, therefore inhibiting fast axonal anterograde and retrograde transport. We coupled the intraplantar injection IL-6 with the injection of vehicle, colchicine (100 μg) or nocodazole (10 μg) [8] in the popliteal fossa of mice. The popliteal fossa is next to the sciatic nerve and can provide a reservoir for drug to bathe the sciatic nerve following appropriate injection. Injection of vehicle into the popliteal fossa had no effect on the initiation or maintenance of IL-6 induced hyperalgesic priming (Figure  2A). However, popliteal fossa injection of either colchicine or nocodazole prevented IL-6-mediated mechanical hypersensitivity and hyperalgesic priming to subsequent PGE₂ treatment (Figure  2A). In order to provide a more direct link to retrograde axonal transport as a critical mediator of IL-6-mediated mechanical hypersensitivity and hyperalgesic priming we used a selective inhibitor for cytoplasmic dynein, ciliobrevin D. Ciliobrevin D blocks dynein-dependent microtubule gliding and ATPase activity while sparing kinesin dependent cellular trafficking therefore selectively affecting retrograde transport [16]. Popliteal fossa injection of ciliobrevin D (10 μg) prevented both IL-6-induced mechanical hypersensitivity and hyperalgesic priming (Figure  2B). To assess whether the effects of nocodozole or ciliobrevin D could be attributable to systemic effects we injected these compounds into the contralateral popliteal fossa at the same dose at the time of IL-6 injection. As opposed to ipsilateral injections, neither compound had an effect on IL-6-induced mechanical hypersensitivity at any time point when given contralaterally (Figure  2C). These results strongly suggest that retrograde axonal transport is necessary for IL-6-mediated nociceptive plasticity.

BDNF is released by nociceptors in response to noxious stimulation and has a profound effect on post-synaptic dorsal horn neurons. BDNF expression shows remarkable plasticity in the DRG following inflammation and BDNF expression in nociceptors is critical for inflammatory injury induced nociceptive plasticity. We have recently shown that IL-6-induced mechanical hypersensitivity and hyperalgesic priming are dependent on BDNF/trkB signaling [14]. Interestingly, IL-6-induced hyperalgesic priming can be disrupted at any time by blocking BDNF/trkB signaling suggesting that IL-6 mediated priming fundamentally changes BDNF/trkB signaling between DRG and dorsal horn neurons. How is this regulated? CREB is known to regulate the expression of BDNF in neurons [17], therefore, we determined whether expression of BDNF is increased following i.pl. injection of IL-6 and whether the blockade of retrograde axonal transport would prevent this effect. We again used popliteal fossa injections of ciliobrevin D at the time of i.pl. injections of IL-6 and measured the expression of BDNF in the dorsal root ganglia 3 hours post-injection (a time point at which CREB retrogradely transported from the paw should reach the DRG). We determined that intraplantar injection of IL-6 enhanced the expression of BDNF in the L4-L6 DRG and that this effect was blocked by popliteal fossa injection of ciliobrevin D (Figure  2D). These results link retrograde axonal transport in the sciatic nerve to IL-6-mediated changes in BDNF expression in the DRG, however the experiment cannot specify CREB as a regulator for the IL-6-mediated effects. To address this, we took advantage of CREB binding to cAMP response element (CRE) consensus sequences in DNA, an event normally linked to CREB’s action as a transcription factor. To disrupt the binding of CREB to CREs we administered intrathecal DNA decoy oligonucleotides that contain either the CRE consensus or mutant sequence [18‐20]. The decoy CRE consensus oligonucleotides prevented the development of i.pl. IL-6-mediated mechanical hypersensitivity and hyperalgesic priming whereas CRE mutant oligonucleotides were without effect (Figure  3A). Moreover, i.t. administration of the CRE consensus decoy oligonucleotides, but not mutant oligonucleotides, prevented IL-6-mediated increases in BDNF expression in DRG 3 hrs after IL-6 treament (Figure  3B). These findings link IL-6-mediated enhanced expression of BNDF in the DRG and IL-6-induced mechanical hypersensitivity and hyperalgesic priming to the retrograde transport of CREB.

Nociceptive plasticity in the dorsal horn is regulated by numerous transcription factors which include CREB [21, 22]. A possible interpretation of the CRE decoy oligonucleotide experiment is blockade of CREB action in dorsal horn neurons. Thus, we aimed to demonstrate that CREB acts upstream of BDNF in DRG neurons. To this end, we co-injected i.t. BDNF, at a dose we previously determined to induce mechanical hypersensitivity and hyperalgesic priming of a similar duration and magnitude to IL-6, with and without CRE consensus decoy oligonucleotides. Unlike experiments where IL-6 was injected into the paw, CRE consensus oligonucleotides failed to alter i.t. BDNF-induced mechanical hypersensitivity or hyperalgesic priming (Figure  3C). Therefore, we conclude that CREB acts upstream of BDNF in the DRG and that CREB is not a key mediator of BDNF-induced nociceptive plasticity in the dorsal horn, at least over this time course.Finally we asked whether the effects of ciliobrevin D might be mediated by effects on nerve conduction. Because IL-6-induced mechanical hypersensitivity develops slowly and there is no acute response to IL-6 we turned to capsaicin in these experiments. Mice received popliteal fossa injections of vehicle, lidocaine (2%) or ciliobrevin D (10 μg) and i.pl. injections of 0.9% capsaicin 5-10 min later. While robust nocifensive responses to capsaicin were observed in the vehicle and ciliobrevin D treated mice, lidocaine, as expected, completely blocked this effect (Figure  4A). We then measured mechanical hypersensitivity in response to capsaicin at 1 and 3 hrs after i.pl. injection. At 1 hr, mechanical hypersensitivity was observed in the vehicle and ciliobrevin D groups but at 3 hrs there was an attenuation of mechanical hypersensitivity in the ciliobrevin D group compared to vehicle (Figure  4B). Mechanical hypersensitivity was completely blocked by lidocaine injection (Figure  4B). These results are consistent with the time course of retrograde transport in IL-6 experiments with ciliobrevin D and demonstrate that ciliobrevin D has no effect on acute pain-related behaviors or mechanical hypersensitivity.

Male ICR mice (Harlan, 20-25 g) were used. All animal procedures were approved by the Institutional Animal Care and Use Committee of The University of Arizona and were in accordance with International Association for the Study of Pain and National Institutes of Health Animal Care guidelines. Animals were placed in acrylic boxes with wire mesh floors and allowed to habituate for 1 hour prior to all behavioral testing. The experimenter making measurements was always blinded to the experimental conditions. IL-6 (0.1 ng) was injected into the plantar surface of the left hindpaw in a volume of 25 μl. BDNF was injected intrathecally (i.t.) in a volume of 5 μl. Popliteal fossa injections were done immediately after intraplantar (i.pl.) injections under brief (<3 min) isoflurane anesthesia in a volume of 50 μl. The exception was i.pl. capsaicin experiments where the popliteal fossa injection was done 5-10 min prior to the capsaicin injection. For i.t. treatments compounds were injected immediately after i.pl. injections under brief (<3 min) isoflurane anesthesia in a volume of 5 μl. PGE₂ (100 ng) was injected i.pl. on day 6 or later in a volume of 25 μl. Mechanical testing was done using the up-down method at time points indicated in the text. Nocifensive behavior induced by capsaicin injection was observed over 15 min in mechanical testing boxes following habituation of the animals.

DRG from ICR mice were excised aseptically and placed in Hank's Buffered Salt Solution (HBSS, Invitrogen, Grand Island, NY) on ice. The ganglia were dissociated enzymatically with collagenase A (1 mg/ml, 25 min, Roche, Indianapolis, IN) and collagenase D (1 mg/ml, Roche, Indianapolis, IN) with papain (30 units/ml) for 20 min at 37°C. To eliminate debris, 70 μm (BD Biosciences, San Jose, CA) cell strainers were used. The dissociated cells were resuspended in DMEM/F12 (Invitrogen, Grand Island, NY) containing 1X pen-strep (Invitrogen, Grand Island, NY), 1X GlutaMax, 3 μg/ml 5-FDU (Sigma, St. Louis, MO), 7 μg/ml uridine (Sigma, St. Louis, MO) and 10% fetal bovine serum (Thermo Hyclone, Rockford, IL). The cells were plated in 6-well plates (BD Biosciences) and incubated at 37°C in a humidified 95% air/5%CO₂ incubator. On day 5 the cells were washed in DMEM/F12 media for 30 min followed by treatment with escalating doses of IL-6 for 15 minutes.

Azidohomoalanine (AHA) is a methionine analogue that cells can incorporate into nascentlly synthesized proteins. The left paws of ICR mice were injected with vehicle or IL-6 (0.1 ng) along with AHA (300 ng, Life technologies) in 25 μl sterile phosphate buffered saline (PBS). After 15 min or 2 hrs sciatic nerves, taken 2 cm from the injection site, were harvested. Protein was extracted from the nerves by ultrasonication, centrifugation at 20000 × g for 15 min and collection of the supernatant. Equal amount of protein was used to immunoprecipitate CREB by incubating the supernatant with 1:50 mouse anti-CREB antibody conjugated to Sepharose beads (cat # 3955, Cell Signalling, Danvers, MA) overnight at 4°C. This was followed by centrifugation to precipitate the beads. The pelleted beads were suspended in Tris-SDS buffer (1% SDS and 50 mM Tris-HCl, pH 8.0), centrifuged and the supernatant was collected. At this stage, the supernatant contains the immunoprecipitated CREB where the nascently synthesized form would have incorporated AHA. AHA was biotinylated using Click-it Protein Analysis Detection Kit (Life technologies, Grand Island, NY) according to the manufacturer’s instructions. The biotinylated nascently synthesized CREB was detected by Western blotting.

Protein was extracted from the cells and tissue in lysis buffer (50 mM Tris HCl, 1% Triton X-100, 150 mM NaCl, and 1 mM EDTA at pH 7.4) containing protease and phosphatase inhibitor mixtures (Sigma, St. Louis, MO) with an ultrasonicator on ice, and cleared of cellular debris and nuclei by centrifugation at 14,000 RCF for 15 min at 4°C. Fifteen micrograms of protein per well were loaded and separated by standard 7.5% or 10% SDS-PAGE. Proteins were transferred to Immobilon-P membranes (Millipore, Billerica, MA) and then blocked with 5% dry milk for 3 h at room temperature. The blots were incubated with primary antibody overnight at 4°C and detected the following day with donkey anti-rabbit antibody conjugated to horseradish peroxidase (Jackson Immunoresearch, West Grove, PA). Signal was detected by ECL on chemiluminescent films. Densitometric analyses were performed with Image J software (NIH, Bethesda, MD).

The anti-CREB rabbit antibodies (cat# 9197) were obtained from Cell Signaling (Danvers, MA). The BDNF antibody was from Sigma (cat# AV41970, St. Louis, MO) and βIII-Tubulin was from Promega (cat# G7121, Madison, WI). The CRE concensus (cat# sc-2504) and mutant (cat# sc-2517) oligonucleotides were from Santa Cruz Biotechnologies (Santa Cruz, CA). Human recombinant IL-6 and BDNF were from R&D Systems. Colchicine and nocodazole were from Tocris Bioscience; Ciliobrevin D was from EMD Millipore and PGE₂ was from Cayman Chemical Company. Capsaicin and lidocaine were from Sigma Aldrich. Stock solutions for colchicine, nocodazole and ciliobrevin D were made in cell culture grade 100% DMSO. BDNF and IL-6 stock solution was made in sterile PBS containing 0.1% BSA and TrkB/Fc stock solution was made in sterile PBS. Capsaicin and PGE₂ stock solutions were made in 100% ethanol. Lidocaine stock solutions were made in sterile saline. All drugs except were diluted to final concentrations in sterile PBS for injection.

Data are shown as means and the standard error of the mean (±SEM) of n = 8 independent cell culture wells, n = 6 tissue samples for in vivo Western blotting, and immunoprecipitation or n = 6 independent animals for behavioral studies. Graph plotting and statistical analysis used GraphPad Prism Version 5.03 (Graph Pad Software, Inc. San Diego, CA, USA). Statistical evaluation was performed by one- or two-way analysis of variance (ANOVA), followed, where appropriate, by Dunnet post-hoc analysis for behavioral and Western blot data. Where only two comparisons were made (Figure  1B and C) student’s t-test was used. The a priori level of significance was considered to be p < 0.05.