Interleukin-17 sensitizes joint nociceptors to mechanical stimuli and contributes to arthritic pain through neuronal interleukin-17 receptors in rodents

Authors


Abstract

Objective

Interleukin-17 (IL-17) is considered a proinflammatory cytokine, but whether neuronal IL-17 receptors contribute to the generation of arthritic pain is unknown. This study was undertaken to explore whether IL-17A acts on neurons, whether it sensitizes joint nociceptors, and whether neutralization of IL-17 is antinociceptive.

Methods

We recorded action potentials from rat joint nociceptors after intraarticular injection of IL-17A. We studied the expression of the IL-17A receptor in the rat dorsal root ganglia (DRG), explored the effect of IL-17A on signaling pathways in cultured rat DRG neurons, and using patch clamp recordings, monitored changes of excitability by IL-17A. We tested whether an antibody to IL-17 influences pain behaviors in mice with antigen-induced arthritis (AIA).

Results

A single injection of IL-17A into the rat knee joint elicited a slowly developing and long-lasting sensitization of nociceptive C fibers of the joint to mechanical stimuli, which was not attenuated by neutralizing tumor necrosis factor α or IL-6. The IL-17A receptor was visualized in most rat DRG neurons, the cell bodies of primary sensory neurons. In isolated and cultured rat DRG neurons, IL-17A caused rapid phosphorylation of protein kinase B and ERK, and it rapidly enhanced excitability. In mice with unilateral AIA in the knee, an antibody against IL-17 improved the guarding score and reduced secondary mechanical hyperalgesia at the ipsilateral paw.

Conclusion

Our findings indicate that IL-17A has the potential to act as a pain mediator by targeting IL-17 receptors in nociceptive neurons, and these receptors are particularly involved in inflammation-evoked mechanical hyperalgesia.

In addition to the proinflammatory cytokines tumor necrosis factor α (TNFα), interleukin-6 (IL-6), and IL-1β, the cytokine IL-17 is considered to be a major mediator of immunity and inflammation in diseases such as rheumatoid arthritis (RA), multiple sclerosis, and others (1–5). IL-17A, the prototype member of the IL-17 family, is secreted from Th17, CD8+ T cells, γ/δ T cells, natural killer cells, and activated monocytes and neutrophils (4). Mechanistically, IL-17A induces the production of mediators of innate and adaptive immunity (2–6). Elevated IL-17A levels have been observed in the synovial fluid of RA patients (4, 5). Reduction of IL-17A through genetic deficiency or antibody blockade ameliorated disease activity in preclinical models of RA, whereas overexpression of IL-17A aggravated collagen-induced arthritis (CIA) (6). The use of monoclonal antibodies against IL-17 in humans effectively reduced the severity of inflammatory diseases (4, 7).

Recent studies have shown that TNFα, IL-6, and IL-1β are also involved in pain generation during arthritis, in part by directly targeting primary nociceptive neurons which express receptors for these cytokines (8–10). Hence, for example, the antinociceptive effect of TNFα neutralization in RA and experimental immune-mediated arthritis results not only from an amelioration of inflammation and/or destruction, but also, at least in part, from the reversal of cytokine-mediated neuronal sensitization of pain pathways (8, 11, 12). The involvement of proinflammatory cytokines in neuropathic pain states further underscores the neuronal targets of cytokines (13–15).

Whether IL-17 also contributes to pain by acting on neuronal targets has not been studied. Because the intraarticular injection of IL-17 caused experimental hyperalgesia in a behavioral study (16), and because IL-17 receptor A (IL-17RA), one of the 5 IL-17 receptors (IL-17RA through IL-17RE) (3), is ubiquitously expressed, including by neurons of the central nervous system (17), in the present study we tested the hypothesis that IL-17A has the potential to contribute to the generation of inflammatory joint pain. Using in vivo and in vitro approaches, we found that IL-17A causes distinct effects in primary afferent neurons which express IL-17RA and that IL-17A is particularly involved in the generation of mechanical hyperalgesia, a major burden of arthritis (11, 12, 18–20).

MATERIALS AND METHODS

Electrophysiologic recordings of rat knee joint afferents in vivo.

All experiments were approved by the Thuringian government commission for animal protection (permit #02-013/08). Adult male Wistar rats (older than 90 days, weighing 300–400 gm; n = 43) (supplied by the Animal Facility of the University Hospital Jena) were anesthetized with 100 mg/kg sodium thiopentone (Altana) intraperitoneally and supplemental doses (20 mg/kg intraperitoneally) as necessary to maintain areflexia. Rats breathed spontaneously, and mean arterial blood pressure and body temperature were controlled. The right thigh was exposed from the knee joint to the groin, a fastener was used to fix the right femur, and the right hind paw was fixed in a shoe-like holder that allowed calibrated rotation of the lower leg at the knee joint. Using platinum wire electrodes, action potentials (APs) were recorded from single fibers of the medial articular nerve, which has receptive fields on the knee joint. The local mechanical threshold was determined with calibrated von Frey hairs (1.6–52 gm), and the conduction velocity of the nerve fibers (≤1.25 meters/second for C fibers, 1.25–10 meters/second for Aδ fibers, and ≥10 meters/second for Aβ fibers) was determined by electrical stimulation of the mechanical receptive field. Mechanical test stimuli, namely, innocuous and noxious outward torque (20 mNm and 40 mNm, respectively; 15 seconds each), were applied before and after injection of the test substance (in 0.1 ml) into the knee joint cavity. Test substances were phosphate buffered saline (PBS), recombinant rat IL-17A (eBioscience), etanercept, and recombinant human soluble gp130 (both from R&D Systems), all dissolved in 1% bovine serum albumin (BSA) in PBS. Changes in responses within groups (before versus after treatment) were analyzed statistically using Wilcoxon's matched pairs signed rank test or paired t-test (if normally distributed).

Localization of IL-17RA–like immunoreactivity in rat dorsal root ganglia (DRG) sections.

DRGs from all spinal segments of adult Wistar rats were fixed in 4% paraformaldehyde in PBS for 24 hours, then watered, dehydrated with a graded ethanol series and xylene (100%) for 10 minutes, transferred to methyl benzoate (100%) overnight, embedded in paraffin (Histosec; Merck), and cut into 5-μm sections that were dewaxed and autoclaved for 15 minutes (120°C, 1 bar) in 0.1 mole/liter citrate buffer (pH 6.0). After incubation with PBS with Triton X-100 and 2% goat serum (Rockland), sections were incubated overnight at 4°C with a primary rabbit polyclonal antibody (1:200 in PBS containing 1% Triton X-100 and 1% gelatin from cold water fish skin). The antibody was raised against amino acids 32–200, mapping within an N-terminal extracellular domain of IL-17RA of human origin (Santa Cruz Biotechnology). For visualization we used a biotinylated goat anti-rabbit antibody (1:200; Dako) (2 hours) and an avidin–biotin–peroxidase complex (Vectastain Elite ABC kit; Vector) (40 minutes). Sections were developed with Jenchrom px blue (MoBiTec), dehydrated, and embedded in Entellan (Merck). In every second section of 12 sections per ganglion, the proportion of neurons with IL-17RA–like immunoreactivity was determined using a light microscope (Axioplan 2; Zeiss) coupled to a CCD video camera and an image analyzing system (KS 300; Zeiss). A neuron was considered to be positively labeled if its relative gray value was greater than that of neurons from control incubations, which were not treated with the anti–IL-17RA antibody. Only neuronal profiles with a visible nucleus were counted.

Polymerase chain reaction (PCR) of IL-17A and IL-17RA.

Total RNA was obtained from rat DRGs or DRG cells cultured for 1–2 days using an RNeasy Plus Mini kit (Qiagen). For reverse transcription we used a RevertAid First-Strand complementary DNA (cDNA) synthesis kit (Fermentas). All messenger RNA templates were transcribed in cDNA using the oligo(dT)18 primer. Primers for IL-17RA were 5′-CTGCCAAAATGACTGCTTGAG-3′ (forward) and 5′-GCCGAGTAGACGATCCAGAC-3′ (reverse) (297 bp), and primers for IL-17A were 5′-GGAGAATTCCATCCATGTGC-3′ (forward) and 5′-CAGAGTCCAGGGTGAAGTGG-3′ (reverse) (227 bp). For PCR (MyCycler Thermal Cycler; Bio-Rad) we used Platinum (Invitrogen) or TrueStart Hot Start (Fermentas) Taq DNA polymerase. The Master Mix consisted of 1× PCR buffer, 2 mM MgCl2, 1–1.5 units Taq DNA polymerase per 50 μl PCR mixture, 0.2 mM dNTP mixture (10 mM dNTP Mix; Fermentas), and 0.5 μM primer. For amplification, 23 μl of Master Mix plus 2 μl cDNA template were used. Reaction conditions were as follows: step 1, 10 minutes at 95°C; step 2, 30 seconds at 95°C; step 3, 20 seconds at 60°C; step 4, 30 seconds at 72°C; and step 5, 3 minutes at 72°C. Steps 2–4 were repeated 30–35 times. PCR mixtures were analyzed with 3% Tris–acetate–EDTA agarose gel containing ethidium bromide. Low Range DNA Ladder (Fermentas) was used as a size standard.

Culturing of rat DRG neurons.

DRGs from all spinal segments of male Wistar rats (age 60 days) were treated with 215 units/ml of type II collagenase (Paesel & Lorei) dissolved in Ham's F-12 medium (Gibco BRL) for 100 minutes at 37°C. Next, ganglia were placed in Dulbecco's modified Eagle's medium (DMEM; Gibco) containing 10,000 units/ml trypsin (Sigma) for 11 minutes at 37°C. Cells were then dispersed with a fire-polished Pasteur pipette and collected by centrifugation (at 500g for 8 minutes). Cell pellets were suspended in Ham's F-12 medium containing 10% heat-inactivated horse serum, 100 units/ml penicillin, 100 μg/ml streptomycin, 1 mmole/liter L-glutamine (all from Gibco), and 10 ng/ml nerve growth factor (Paesel & Lorei), plated on coverslips coated with poly-L-lysine (50 μg/ml; Sigma) or 12-well plates (Nunc) at 37°C in a humidified incubator gassed with 5% CO2 and air. Cells were fed with culture medium and used within 30 hours. For stimulation experiments, cells were starved for 4 hours in pure DMEM.

Stimulation experiments, sodium dodecyl sulfate–polyacrylamide gel electrophoresis (PAGE), and Western blotting.

DRG cells were treated with 100 ng/ml of recombinant rat IL-17A (ImmunoTools) in DMEM for 1–60 minutes. The reaction was stopped on ice, medium was aspirated, and cells were scraped and harvested with lysis buffer (20 mmoles/liter HEPES, 10 mmoles/liter EGTA, 40 mmoles/liter β-glycerophosphate, 2.5 mmoles/liter MgCl2, and 1% Nonidet P40, pH 7.4, supplemented with 1 mmole/liter phenylmethylsulfonyl fluoride, 2 mmoles/liter Na3VO4, and 2 μg/ml leupeptin). After a freeze–thaw cycle, cell lysates were centrifuged and 6× Laemmli loading buffer was added to supernatants. Three independent preparations were produced.

Proteins were separated on 12.5% PAGE gels at 125V and transferred to a PVDF membrane (Millipore). Immunoblotting was then performed with rabbit anti-phospho–protein kinase B (PKB)/Akt monoclonal antibody against the residue Ser473 (pPKB/Akt, 1:1,000), rabbit anti-PKB/Akt polyclonal antibody (1:1,000), mouse anti–phospho–ERK-1/2 monoclonal antibody against the residues Thr202 and Tyr204 (pERK, 1:2,000), and rabbit anti–β-actin polyclonal antibody (1:1,000) (all from Cell Signaling Technology). For visualization, we used horseradish peroxidase–linked secondary antibodies according to the recommendations of the manufacturer (KPL) with enhanced chemiluminescence reaction using SuperSignal West Dura extended duration substrate (Thermo Scientific) and a CCD camera system (Synoptics).

Patch clamp recordings of cultured rat DRG neurons.

APs from small and medium-sized DRG neurons (18–35 μm; cell bodies of C and Aδ fibers) that were cultured for 12–30 hours were recorded in the current-clamp mode. The bath was perfused with HEPES solution (control; 140 mmoles/liter NaCl, 5 mmoles/liter KCl, 2 mmoles/liter CaCl2, 2 mmoles/liter MgCl2, 10 mmoles/liter glucose, and 10 mmoles/liter HEPES, pH 7.4). Test compounds were added with an application system. The recording pipettes contained 140 mmoles/liter KCl, 10 mmoles/liter NaCl, 1 mmole/liter MgCl2, 0.5 mmoles/liter CaCl2, 2 mmoles/liter Na2-ATP, 5 mmoles/liter EGTA, 10 mmoles/liter HEPES, and 10 mmoles/liter sucrose, pH 7.2. We included only neurons with a membrane potential less than −45 mV. To assess neuronal excitability, APs were elicited by current injection through the recording pipette. At the resting potential, current was applied at amplitudes of 10–70 pA (10 pA steps, pulse duration 5 msec, interpulse interval 2 seconds) until an AP with the typical overshoot was elicited. This protocol was repeated every 2 minutes before application of IL-17A, and within 5–7 minutes after application of IL-17A to the bath. In addition, ramp current (0 pA to 3× threshold current) was applied for 240 msec, and the latency of the first AP and the numbers of elicited APs were measured before and during IL-17A application. Only HEPES was used in control recordings.

Treatment of mice with antigen-induced arthritis (AIA) with antibody against IL-17.

All experiments were approved by the Thuringian government commission for animal protection (permit #02-013/08). C57Bl/6 mice (7–8 weeks old; Charles River) were immunized 21 and 14 days before AIA induction with a subcutaneous injection of 100 μg of methylated BSA (mBSA; Sigma) emulsified with 50 μl of Freund's complete adjuvant (Sigma), which was supplemented with 2 mg/ml Mycobacterium tuberculosis H37Ra (Difco), and 5 × 108 heat-inactivated Bordetella pertussis germs (Chiron-Behring) were applied intraperitoneally. For induction of monarticular AIA, 100 μg of mBSA in 25 μl of 0.9% NaCl was injected into the right knee joint cavity (on day 0).

Mice (n = 9) received an intraperitoneal injection (100 μg) of a rat monoclonal antibody against murine IL-17 (MAB421; R&D Systems) 3 days before as well as 5 hours before AIA induction. This antibody was effective in other studies of arthritis, including studies of AIA (21–23). An additional 9 mice received rat monoclonal IgG2a isotype control (eBioscience) in the same manner.

Swelling at the knee joint was assessed by measuring the mediolateral joint diameter using Oditest Vernier calipers (Kroeplin). Gait abnormalities of the ipsilateral hind limb were scored on a scale of 0–5, where 0 = normal walking, 1 = slight limping, 2 = persistent severe limping (still touching the floor), 3 = severe limping with partial guarding of the ipsilateral hind limb (sometimes not touching the floor), 4 = mainly guarding of the ipsilateral hind limb (most times not touching the floor), and 5 = no walking at all. The mechanical pain thresholds at the hind paws (secondary mechanical hyperalgesia) were determined using a dynamic plantar esthesiometer (Ugo Basile), which applied increasing pressure (increase rate 1 gm/second; cutoff value 10 gm) to the plantar surface of the paw. The weight force (in grams) needed to elicit leg withdrawal was averaged from 3 consecutive stimuli. Two tests were performed during the immunization phase to establish baseline. For statistical analysis we used repeated-measures analysis of variance followed by post hoc Student's 2-tailed t-tests. Differences within groups were tested using paired t-tests.

RESULTS

Sensitization of rat joint nociceptors to mechanical stimuli by IL-17A injection.

An important neuronal mechanism of mechanical hyperalgesia in joints is the sensitization of joint nociceptors to mechanical stimuli, called mechanical sensitization (24). We investigated whether the injection of IL-17A into the normal rat knee joint causes mechanical sensitization, i.e., an increase in the responses of joint nociceptors to the rotation of the joint. The peristimulus time histograms in Figure 1A show an increase in responses (more APs) of 2 C fibers to innocuous and noxious outward rotation 3 hours after a single intraarticular injection of IL-17A.

Figure 1.

Long-term effect of the intraarticular injection of interleukin-17A (IL-17A) on the responses of nociceptive C fibers and Aδ fibers of the rat knee joint to mechanical stimuli. A, Peristimulus time histograms showing responses of 2 representative C fibers to innocuous (Innoc.) outward rotation (20 mNm, 15 seconds) and noxious (Nox.) outward rotation (40 mNm, 15 seconds) before and 3 hours after injection of IL-17A. R = period for recording ongoing discharges (60 seconds). B, Effect of IL-17A on the responses of C fibers to innocuous rotation. The mean ± SEM baseline values before IL-17A injection were 45 ± 25 action potentials (APs)/15 seconds in the 5 ng group, 17 ± 7 APs/15 seconds in the 50 ng group, and 25 ± 14 APs/15 seconds in the 100 ng group. C, Effect of IL-17A on responses of C fibers to noxious rotation. The mean ± SEM baseline values were 97 ± 38 APs/15 seconds in the 5 ng group, 78 ± 28 APs/15 seconds in the 50 ng group, and 172 ± 42 APs/15 seconds in the 100 ng group. D, Effect of IL-17A on responses of Aδ fibers to noxious rotation. The mean ± SEM baseline values were 299 ± 42 APs/15 seconds in the 5 ng group, 193 ± 51 APs/15 seconds in the 50 ng group, 112 ± 47 APs/15 seconds in the 100 ng group, and 303 ± 60 APs/15 seconds in the 500 ng group. The baseline value before IL-17A injection was set to 0. Values in B–D are the mean ± SEM. ∗ = P < 0.05 at the indicated time point and at all later time points (arrows) compared to baseline, by Wilcoxon's matched pairs signed rank test.

The effects of 3 different doses of intraarticular IL-17A on the responses of 3 groups of C fibers to innocuous and noxious outward rotation of the joint are shown in Figures 1B and C. Responses to innocuous outward rotation (Figure 1B) did not change after treatment with 5 ng of IL-17A, showed a small and nonsignificant increase after treatment with 50 ng of IL-17A, and showed a significant increase after treatment with 100 ng of IL-17A (P < 0.05 by Wilcoxon's matched pairs signed rank test). Responses to noxious outward rotation (Figure 1C) increased significantly compared to baseline after treatment with 50 ng and 100 ng of IL-17A (P < 0.05 by Wilcoxon's matched pairs signed rank test). In addition, the local mechanical (von Frey) threshold in the receptive fields dropped significantly after treatment with 50 ng IL-17A (from a mean ± SEM of 24.4 ± 4.3 gm to 10.8 ± 2.7 gm; n = 6 [P < 0.01 by paired t-test]) and after treatment with 100 ng IL-17A (from 26.2 ± 5.3 gm to 18.0 ± 4.5 gm; n = 7) [P < 0.05 by paired t-test]), but it remained unchanged after treatment with 5 ng of IL-17A (24.1 ± 4.5 gm before and 24.5 ± 5.7 gm after IL-17A treatment; n = 7). In contrast, 500 ng of IL-17A reduced the responses of C fibers and deteriorated the fibers within 3 hours.

In thin myelinated Aδ fibers, the intraarticular injection of 50 ng or 100 ng of IL-17A altered neither the responses to rotation of the joint nor the local mechanical thresholds. After treatment with 500 ng IL-17A (a dose which reduces C fiber responses, as described above), the responses of Aδ fibers were significantly increased. However, IL-17A at 5 ng caused a pronounced and significant reduction in the responses of Aδ fibers to both innocuous and noxious outward rotation of the joint (Figure 1D). Vehicle alone did not significantly alter the responses within 3 hours (n = 6 fibers).

Because IL-17A may induce the production of IL-6 and TNFα and because, e.g., TNFα and IL-17 can act synergistically on several cell types (25, 26), we also assessed mechanical sensitization by IL-17A in the presence of etanercept, which blocks the sensitization of C fibers by TNFα (10), and in the presence of soluble gp130, which blocks the sensitization of C fibers by IL-6 (9). The increase in responses at 3 hours was similar in the 3 groups (IL-17A alone, IL-17A plus etanercept, and IL-17A plus soluble gp130) (Figure 2). Thus, these results indicate that IL-17A can sensitize C fibers independent of the effects of TNFα or IL-6.

Figure 2.

Neither etanercept nor soluble gp130 (sgp130) inhibits interleukin-17A (IL-17A)–induced mechanical sensitization. Increases in the responses of C fibers to innocuous (Innoc.) and noxious (Nox.) outward rotation (OR) of the rat knee within 3 hours after intraarticular injection of either IL-17A alone (n = 7 fibers), IL-17A plus etanercept (n = 8 fibers), or IL-17A plus soluble gp130 (n = 6 fibers) are shown. The increases were similar in all treatment groups, and all increases were significant compared to the preinjection baseline (by Wilcoxon's matched pairs signed rank test). The mean ± SEM baseline response values before injection for innocuous outward rotation were 25 ± 14 action potentials (APs)/15 seconds in the IL-17A group, 43 ± 28 APs/15 seconds in the IL-17A plus etanercept group, and 26 ± 16 APs/15 seconds in the IL-17A plus soluble gp130 group. The baseline response values before injection for noxious outward rotation were 172 ± 42 APs/15 seconds in the IL-17A group, 197 ± 61 APs/15 seconds in the IL-17A plus etanercept group, and 150 ± 37 APs/15 seconds in the IL-17A plus soluble gp130 group. Data are shown as box plots, where the boxes represent the 25th to 75th percentiles, the lines within the boxes represent the median, and the error bars represent the 5th and 95th percentiles. Circles indicate outliers.

Expression of IL-17RA in rat DRG neurons.

In order to explore the expression of IL-17RA in primary afferent neurons, we performed immunohistochemistry and PCR on rat DRG neurons, the cell bodies of primary afferent fibers. Anti–IL-17RA antibody labeled somata of small, medium-sized, and large neurons and satellite cells in rat DRG sections (Figures 3A and B). A mean ± SEM of 86 ± 6% of the neurons were labeled. Axons in peripheral DRG processes were also stained. PCR of material from either whole DRGs (Figure 3C) or cultured DRG neurons (Figure 3D) revealed PCR products of IL-17RA (297 bp). No PCR product of IL-17A (227 bp) was identified in cultured DRG neurons (Figure 3E).

Figure 3.

Localization of interleukin-17 receptor A (IL-17RA) in rat dorsal root ganglia (DRG) neurons. A, IL-17RA–like immunoreactivity in a rat DRG section. Asterisk indicates a typical anti–IL-17RA antibody–labeled neuron; plus sign indicates a typical nonlabeled neuron; greater than sign indicates a positively labeled satellite cell. Boxed area shows a labeled neuron (n) and a satellite cell (s; arrow). Bar = 15 μm. B, Size distribution of DRG neurons and proportions of neurons with IL-17RA–like immunoreactivity (IR; n = 5 rats). C, Gel electrophoresis of IL-17RA polymerase chain reaction (PCR) products (297 bp) of DRGs from 4 rats. D, Gel electrophoresis of IL-17RA PCR products (297 bp) of rat DRG neurons from 4 cultures. E, Lack of IL-17A PCR product (227 bp) in rat DRG neurons from 4 cultures. Values in C, D, and E are the basepairs of the size-standard DNA used.

IL-17A acts on isolated and cultured rat DRG neurons.

In order to further explore whether IL-17A acts directly on rat DRG neurons, we exposed cultured rat DRG neurons to IL-17A and determined whether IL-17A activates typical IL-17A–activated signaling pathways (26, 27) in these neurons. Western blot analysis showed increased phosphorylation of PKB/Akt and ERK-1/2 within 5 minutes after the exposure of neurons to IL-17A (100 ng/ml) (Figure 4). Thus, IL-17A activates neurons rapidly.

Figure 4.

Activation of signaling kinases in cultured rat dorsal root ganglia neurons by interleukin-17A (IL-17A). Western blot analysis showed an increase in phosphorylated protein kinase B (pPKB)/Akt and pERK-1/2 within 5 minutes after the exposure of neurons to IL-17A (100 ng/ml). The control incubation (indicated by the minus sign) lasted 60 minutes.

Using patch clamp recordings of isolated cultured rat DRG neurons, we tested whether IL-17A renders neurons more excitable. We used 2 different protocols to measure the threshold for the elicitation of APs in small and medium-sized rat DRG neurons upon current injection (Figure 5). Figure 5A shows that the current that was necessary for the elicitation of an AP in a rat DRG neuron dropped from 40 pA to 30 pA after application of 100 ng/ml IL-17A to the bath. Figure 5B shows the results for all neurons tested in this way. Of the 23 control neurons tested with buffer only (0 ng/ml IL-17A), only 2 showed a decrease in threshold during the testing period, whereas the threshold remained unchanged or even increased in the other neurons. Application of 10 ng/ml IL-17A to the bath did not significantly change this pattern (Figure 5B). After application of 100 ng/ml IL-17A, 9 of 14 neurons showed a decrease in threshold, and this proportion was significantly higher than the proportion of control neurons that showed a decrease.

Figure 5.

Increase in the excitability of isolated rat dorsal root ganglia (DRG) neurons by interleukin-17A (IL-17A). A, Reduction of the threshold from 40 pA to 30 pA by IL-17A in a rat DRG neuron, assessed by the application of single current pulses of 5 msec with increasing amplitudes. B, Application of the protocol shown in A in 23 control rat DRG neurons (0 ng/ml), in 13 rat DRG neurons in which 10 ng/ml IL-17A was applied to the bath, and in 14 rat DRG neurons in which 100 ng/ml IL-17A was applied to the bath. The lines show the initial thresholds and thresholds 5–7 minutes after the application of HEPES (0 ng/ml IL-17A) or IL-17A. After application of 100 ng/ml IL-17A, a higher proportion of the neurons showed a decrease in threshold. ∗∗ = P < 0.01 by chi-square test. C, Reduction in the threshold for elicitation of action potentials (APs; shorter latency) and increase in the number of APs in a rat DRG neuron (same neuron as in A) by IL-17A, assessed with a ramp-shaped current injection protocol. D, Reduction in latency and increase in the number of APs by IL-17A (100 ng/ml) in neurons tested with the ramp-shaped current protocol. Data are shown as box plots, where the boxes represent the 25th to 75th percentiles, the lines within the boxes represent the median, and the error bars represent the 5th and 95th percentiles. Circles indicate outliers. ∗ = P < 0.05; ∗∗ = P < 0.01, by Wilcoxon's matched pairs signed rank test.

These 14 neurons were also tested using a ramp-shaped current injection protocol. After application of IL-17A to the bath, the neuron subjected to the protocol illustrated in Figure 5C showed the first AP at a shorter latency (and hence at a lower current amplitude), and it fired more APs during the current ramp. In the entire group of 14 neurons subjected to application of 100 mg/ml IL-17A, latency was significantly shorter (indicating a reduction in threshold) and the number of APs was significantly higher compared to control neurons (Figure 5D). These findings in isolated neurons show that IL-17A has the potential to directly increase the excitability of primary afferent neurons.

Reduction in nociception in murine AIA by neutralization of IL-17.

Finally, we tested whether neutralization of IL-17 influences mechanical hyperalgesia in arthritis. Because no antibodies were available for rats we used mice. We induced unilateral AIA in the right knee joint of wild-type mice and treated one group with an antibody against IL-17 and one group with rat IgG2a isotype control. AIA critically depends on T cells, and Th17 cells are involved (28, 29). Injection of the antibody twice (3 days and 5 hours before the induction of AIA) attenuated joint swelling, but this effect did not reach statistical significance (Figure 6A). However, treatment with anti–IL-17 antibody significantly reduced the guarding score in the inflamed leg (Figure 6B). It also reduced secondary mechanical hyperalgesia at the paw. The control mice with AIA (Figure 6C, top) showed a reduction in mechanical threshold at the ipsilateral paw and a slight increase in threshold at the contralateral paw on days 1 and 3. (The increase in threshold at the uninflamed paw is most likely a counterregulatory reaction to prevent load on the inflamed side.) Such a difference was not observed in antibody-treated mice (Figure 6C, bottom).

Figure 6.

Results of behavioral experiments in mice, showing the involvement of interleukin-17 (IL-17) in mechanical hyperalgesia. A, Joint swelling in mice with antigen-induced arthritis (AIA). Swelling was slightly but not significantly reduced in anti–IL-17–treated mice (F[1,14] = 2.266; P = 0.154). B, Significant reduction in the guarding score for the leg with the inflamed knee in mice with AIA treated with anti–IL-17 antibody (F[1,14] = 14.324; P = 0.002). C, Reduction in mechanical threshold at the paw ipsilateral to the inflamed knee, slight increase in threshold at the paw contralateral to the inflamed knee (top) (F[1,16] = 6.035; P = 0.026), and prevention of these changes by anti–IL-17 antibody (bottom) (F[1,16] = 0.162; P = 0.693). Values are the mean ± SEM (n = 9). ∗ = P < 0.05; ∗∗ = P < 0.01, by t-test (comparison of values at a single time point). # = P < 0.05 at this time point and all later time points versus preinflammatory baseline, by matched pairs t-test (intragroup comparison).

DISCUSSION

The present study shows that IL-17A has the potential to act as a pain mediator, in addition to its immunologic functions. IL-17A induces a long-term sensitization of rat joint nociceptors to mechanical stimuli, and this sensitization is not blocked by TNFα neutralization or IL-6 neutralization. In murine AIA, neutralization of IL-17 improved the guarding score and secondary mechanical hyperalgesia. Primary sensory neurons express IL-17RA. Application of IL-17A to isolated rat sensory neurons activated signaling cascades in the neurons and enhanced their excitability. Thus, IL-17A plays a role in mechanonociception, and primary sensory neurons themselves are a target of IL-17A.

Similar to TNFα and IL-6 (9, 10), a single injection of IL-17A into the normal rat knee joint caused a slowly developing and persistent increase in the responses of nociceptive C fibers to both innocuous and noxious rotation of the joint. The time course of sensitization was dose dependent (doses were in the same range as in the behavioral experiments of Pinto et al [16] and McNamee et al [30] in mice). Because the IL-17A–induced sensitization was not prevented by etanercept or by soluble gp130, we conclude that IL-17A can sensitize joint afferents by itself. This does not exclude synergistic actions of cytokines under pathologic conditions (30).

Interestingly, nociceptive Aδ fibers were only sensitized at a dose of IL-17A that was deteriorating for C fibers, suggesting that IL-17A is mainly a sensitizer of C fibers. Surprisingly, the lowest dose of IL-17A used (subthreshold for the sensitization of C fibers) reduced the responses of Aδ fibers. Since both Aδ and C fiber inputs determine the responses of spinal cord neurons to peripheral stimulation, subthreshold doses of IL-17A for C fibers may even reduce the input from the periphery by inhibiting Aδ fiber responses and thus inhibit spinal nociceptive processing.

Several findings provide evidence that the effects of IL-17A on mechanosensitivity can result from direct actions of IL-17A on neurons. First, both immunohistochemistry and PCR revealed the expression of IL-17RA in rat DRG neurons. Notably, the vast majority of rat DRG neurons expressed IL-17RA. Thus, IL-17RA is more widely expressed in rat DRG neurons than are TNF receptors (8) and IL-1 receptors (31), which are only localized in ∼30–50% of the DRG neurons. Second, exposure of isolated DRG neurons to IL-17A evoked phosphorylation of PKB/Akt and ERK within 5 minutes. In particular, the up-regulation of pERK is usually observed when DRG neurons are rendered hyperexcitable (32). Third, patch clamp recordings showed the generation of hyperexcitability of small to medium-sized DRG neurons within a few minutes after application of IL-17A to the bath. Thus, IL-17A regulates the activation of voltage-gated ion channels. Sensitizing mediators, such as prostaglandins (33) and TNFα (34), sensitize not only ion channels for the transduction of physical stimuli, but also voltage-gated ion channels, such as tetrodotoxin-resistant sodium channels, which are involved in the generation of APs (33–36). Preliminary data indicate that sodium channels such as Nav1.8 are enhanced by IL-17A (Natura and Schaible: unpublished observations).

In several previous studies, neutralization of IL-17 was found to attenuate inflammation and destruction in CIA (37, 38). In the present study, we observed a weak reduction in swelling in mice with AIA treated with anti–IL-17 compared to untreated mice with AIA, which did not, however, reach significance. However, the guarding score was significantly improved by IL-17 neutralization. Furthermore, the disappearance of the paw withdrawal asymmetry upon mechanical testing in antibody-treated mice indicates that IL-17 is also involved in spinally mediated secondary mechanical hyperalgesia that is remote from the inflamed region. Taken together, these findings and the data from the recordings of rat joint nociceptors show a role of IL-17A in mechanonociception. In human RA, a fully human antibody to IL-17A (AIN457) reduced inflammation and significantly decreased the number of joints with pressure pain as early as 1 week and up to 16 weeks after 1 or 2 injections of the antibody (7), suggesting a role of IL-17 in human mechanical hyperalgesia.

In summary, the present study provides evidence that primary nociceptive neurons are a direct target of IL-17A and that IL-17A plays a role in the generation and maintenance of mechanical hyperalgesia. Although IL-17A exerts these neuronal effects on its own, it remains to be explored whether this cytokine acts synergistically with other cytokines on neurons, whether there is a hierarchy among cytokines concerning the neuronal effects, or whether each cytokine contributes a specific neuronal effect. Since IL-17RA is also expressed in satellite cells of DRGs, IL-17A may also be involved in neuropathic pain because satellite cells play a role in the pathologic destructive and repair processes in the DRGs in the aftermath of neuronal damage (39).

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Schaible had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Richter, Natura, Ebbinghaus, Segond von Banchet, Hensellek, König, Bräuer, Schaible.

Acquisition of data. Richter, Natura, Ebbinghaus, Segond von Banchet, Hensellek, König, Bräuer, Schaible.

Analysis and interpretation of data. Richter, Natura, Ebbinghaus, Segond von Banchet, Hensellek, König, Bräuer, Schaible.

Acknowledgements

The authors thank Mrs. Antje Wallner and Mrs. Konstanze Ernst for technical assistance.

Ancillary