Methylprednisolone prevents nerve injury-induced hyperalgesia in neprilysin knockout mice

Ò

PAIN 155 (2014) 574–580

www.elsevier.com/locate/pain

Methylprednisolone prevents nerve injury-induced hyperalgesia in neprilysin knockout mice Lan He a,2, Nurcan Üçeyler a, Heidrun H. Krämer b,3, Maria Nandini Colaço a, Bao Lu c, Frank Birklein b,1, Claudia Sommer b,⇑,1 a b c

Department of Neurology, University Hospital of Würzburg, Würzburg, Germany Department of Neurology, Johannes Gutenberg University Mainz, Mainz, Germany Boston Children’s Hospital, Harvard Medical School, Boston, MA, USA

Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.

a r t i c l e

i n f o

Article history: Received 1 February 2013 Received in revised form 16 November 2013 Accepted 4 December 2013

Keywords: Complex regional pain syndrome Endothelin-1 Neprilysin Nerve injury Neutral endopeptidase

a b s t r a c t The pathophysiology of the complex regional pain syndrome involves enhanced neurogenic inflammation mediated by neuropeptides. Neutral endopeptidase (neprilysin, NEP) is a key enzyme in neuropeptide catabolism. Our previous work revealed that NEP knock out (ko) mice develop more severe hypersensitivity to thermal and mechanical stimuli after chronic constriction injury (CCI) of the sciatic nerve than wild-type (wt) mice. Because treatment with glucocorticoids is effective in early complex regional pain syndrome, we investigated whether methylprednisolone (MP) reduces pain and sciatic nerve neuropeptide content in NEP ko and wt mice with nerve injury. After CCI, NEP ko mice developed more severe thermal and mechanical hypersensitivity and hind paw edema than wt mice, confirming previous findings. Hypersensitivity was prevented by MP treatment in NEP ko but not in wt mice. MP treatment had no effect on protein levels of calcitonin-gene related peptide, substance P, and bradykinin in sciatic nerves of NEP ko mice. Endothelin-1 (ET-1) levels were higher in naïve and nerve-injured NEP ko than in wt mice, without an effect of MP treatment. Gene expression of the ET-1 receptors ETAR and ETBR was not different between genotypes and was not altered after CCI, but was increased after additional MP treatment. The ETBR agonist IRL-1620 was analgesic in NEP ko mice after CCI, and the ETBR antagonist BQ-788 showed a trend to reduce the analgesic effect of MP. The results provide evidence that MP reduces CCI-induced hyperalgesia in NEP ko mice, and that this may be related to ET-1 via analgesic actions of ETBR. Ó 2013 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction Studies on the early changes in complex regional pain syndrome (CRPS) underscore the importance of neuropeptides. The nonapeptide bradykinin, which was found increased systemically in CRPS patients [3], is a potent algogenic mediator and is involved in edema formation [16]. In the warm and red ‘‘inflammatory phenotype’’ of CRPS, electrical stimulation of primary afferents results in an intense neurogenic flare and edema. The former is supposed to be mediated by calcitonin gene-related peptide (CGRP), the

⇑ Corresponding author. Address: Department of Neurology, University Hospital of Würzburg, Josef-Schneider-Str. 11, D-97080 Würzburg, Germany. Tel.: +49 931 201 23763; fax: +49 931 201 23697. E-mail address: [email protected] (C. Sommer). 1 Frank Birklein and Claudia Sommer have shared senior authorship. 2 Current address: Department of Neurology, University of Texas, Houston, TX, USA. 3 Current address: Department of Neurology, Justus Liebig University of Gießen, Gießen, Germany.

latter by substance P (SP) [46]. Endothelin-1 (ET-1), a potent vasoconstrictive peptide, is increased in blister fluids of patients with cold and bluish, noninflammatory CRPS extremities [18]. ET-1 is first translated as pre-pro-ET-1 and is cleaved to the active peptide ET-1 by the endothelin converting enzyme-1 (ECE-1). Exogenously administered ET-1 elicits pain in humans and experimental animals, mostly through ETA receptors (ETAR) [1]. ETB receptors (ETBR) may have an analgesic effect, depending on doses and local conditions, as demonstrated in an animal model of CRPS [34]. Because of the still-incomplete understanding of the pathophysiology of CRPS and the lack of large-scale clinical trials, treatment of CRPS remains mostly empirical [31,35]. Systemic glucocorticoids seem to be effective when administered early in the course of CRPS [4,11]. Consistent with this, continuously infused methylprednisolone (MP) reversed hind paw edema and protein extravasation in rats with tibial fracture [27] and also reversed thermal and mechanical hyperalgesia in sciatic nervetransected rats [26].

0304-3959/$36.00 Ó 2013 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.pain.2013.12.003

Ò

L. He et al. / PAIN 155 (2014) 574–580

Human neutral endopeptidase (neprilysin, NEP), a zinc endopeptidase, degrades several peptides, including CGRP, SP, and ET-1, and regulates osteoblasts [21], which might be involved in CRPS pathogenesis [36]. We previously demonstrated that after chronic constriction injury (CCI) of the sciatic nerve, mice deficient of NEP (NEP knockout [ko] mice) were more sensitive to heat and mechanical stimuli than wild-type (wt) mice, and developed edema and changes in limb temperature resembling human CRPS II (ie, CRPS with major nerve lesion) [28]. There is no perfect animal model for CRPS, and although CCI is considered a nerve injury model, we took advantage of the CRPS-like phenotype of NEP ko mice after CCI to investigate steroid responsiveness. We hypothesized that a glucocorticosteroid such as MP, which is reportedly effective in the treatment of CRPS [4,11,23], would have analgesic effects in NEP ko mice after CCI and that the endothelin system and its receptors are the basis for this analgesic effect. 2. Materials and methods 2.1. Animals, surgery, and drug treatment NEP ko mice on C57Bl/6 background (>10 generations backcrossed) were generated in the laboratory of B.L. at Boston Children’s Hospital, Harvard Medical School [30] and were bred and maintained at the animal facilities of the University of Würzburg. C57Bl/6 wt mice were purchased from Harlan Winkelmann, Germany. The mice were age-matched (wt mice: 12 weeks ± 0.4; ko mice: 13 weeks ± 0.3) and cage-matched (ie, kept in the same cages and environment). The experiments were approved by the Bavarian State authorities. CCI was performed according to the method described by Bennett and Xie [2] modified for mice [43]. Briefly, animals were anesthetized by isoflurane, and the right sciatic nerve was exposed at the level of the mid-thigh just proximal to its trifurcation. Three ligatures were loosely tied around the nerve at 1 mm distance using 7-0 Prolene suture material (Ethicon, Germany). MP sodium succinate (Upjohn Company, Kalamazoo, MI) was diluted in normal saline (NS) for infusions. At the time of CCI, all mice were implanted with infusion pumps (ALZET 1007D; Alza, Palo Alto, CA). NEP ko and wt mice received either NS or MP (3 mg/kg body weight/day) for 7 days. An additional group of NEP ko and wt mice received MP without CCI (n = 5 to 6 for all groups). To investigate the role of the endothelin receptors, we applied the ETBR agonist IRL-1620 and the ETBR antagonist BQ-788 and assessed pain behavior. Twenty NEP ko mice each received CCI of the right sciatic nerve. Withdrawal thresholds to heat and withdrawal latencies to mechanical stimulation were recorded before CCI and at day 7 after surgery. At day 7, 5 mice received an intraplantar injection of 50 pmol of the ETBR agonist IRL-1620 in a volume of 10 lL (Tocris Bioscience, Wiesbaden-Nordenstadt, Germany), 5 mice received an equal volume of distilled water; another 5 mice were implanted with infusion pumps (ALZET 1007D) with MP (3 mg/kg body weight/day) for 7 days and received an intraplantar injection of 60 nmol of the ETBR antagonist BQ-788 in a volume of 10 lL (Tocris Bioscience, Wiesbaden-Nordenstadt, Germany) on day 7. Five mice received an equal volume of the solvent (ethanol). Behavioral tests were repeated 30 minutes afterward. Injection dosage and time point for testing was adjusted to published data [34]. Mice were killed in deep isoflurane anesthesia, and skin from the footpads of all mice from the agonist and antagonist experiments was obtained for gene expression analysis (see later). 2.2. Behavioral testing For all behavioral tests, the investigator was unaware of the animal genotype.

575

Thermal withdrawal latencies were measured to monitor sensitivity to heat [2]. The mice were tested for paw withdrawal latencies to a thermal stimulus on 3 consecutive days before surgery to habituate animals to the testing device and to obtain baseline values. Mice were then tested at days 3 and 7 after CCI. The animals were put in Plexiglas cages on a glass plate and rested for 30 minutes before the experiment started. A radiant heat source (Plantar Tester, Ugo Basile, Comerio, Italy) was focused on the plantar surface of the hind paw, and the latency from the initiation of the radiant heat until paw withdrawal (paw withdrawal latency) was measured automatically. A maximal cutoff of 15 seconds was used to prevent tissue damage. Each paw was tested 3 times, and the mean withdrawal latency was calculated. The interval between 2 trials on the same paw was at least 5 minutes. Withdrawal thresholds to mechanical stimuli were assessed with von Frey hairs using the up-and-down method [7] on 3 consecutive days before surgery and on days 3 and 7 after surgery. Mice were placed on a wire mesh in Plexiglas cages. The plantar surface was touched perpendicularly with a von Frey hair until the filament was slightly bent. Testing started with the 0.84 mN von Frey hair. If the mouse responded to the touch within 3 seconds by brisk withdrawal of the respective hind paw, the response was interpreted as positive. In case of a positive response, the next weaker stimulus was applied. In case of a negative response, the next stronger stimulus was used. This procedure was performed until 6 responses were recorded. The 50% threshold was calculated using the formula: 50% threshold (mN) = (10[Xf + jd])/10,000  9.8 indicating the force at which individual mice withdrew the hind paw in 50% of trials (Xf = value [log units] of the final von Frey hair used; j = tabular value according to Dixon [15]; d = mean of difference [log units] between stimuli). Thickness of the hind paws was assessed 1 day before and at days 1, 3, 5, and 7 after surgery using a caliper. 2.3. Protein extraction and enzyme-linked immunosorbent assay (ELISA) Wt and NEP ko mice were killed on day 8 after surgery or after initiation of MP treatment, respectively, with an isoflurane overdose, and sciatic nerves were dissected bilaterally. Tissue was immediately frozen in liquid nitrogen and stored at 80 °C until analysis. Samples for ELISA studies were homogenized in proteaseinhibitor-containing phosphate-buffered saline (PBS with aprotinin, leupeptin, pepstatin; Boehringer Mannheim, Germany, pH 7.4) at 39,000 rpm for 30 seconds using a Miccra D-8 power homogenizer (ART, Müllheim-Hügelheim, Germany). After centrifugation for 10 minutes at +4 °C, the supernatant was removed, aliquoted, and assayed in duplicate by a commercial CGRP (Spi-Bio, Paris, France), SP (Spi-Bio), bradykinin (Immundiagnostic AG, Bensheim, Germany), and ET-1 (Cayman, Ann Arbor, MI) enzyme immunoassay according to the manufacturer’s instructions. The assay detection limits were: CGRP: 5 pg/mL; SP 4 pg/ mL; bradykinin 173 pg/mL; ET-1 18 pg/mL. 2.4. Gene expression analysis RNA extraction from sciatic nerve and from skin was performed as described earlier [42,45]. In brief, total RNA from sciatic nerve samples was extracted following the method of Chomczynski with modifications [10]. Frozen nervous tissue was incubated in TRIzol reagent (Invitrogen, Karlsruhe, Germany) and homogenized (Polytron PT 1600E, Kinematica, Luzern, Switzerland). Chloroform was added, and the samples were centrifuged (13,000 rpm, 4 °C, 15 minutes). The upper phase was mixed with glycogen and propanol, and after incubation overnight at 20 °C the samples were

576

Ò

L. He et al. / PAIN 155 (2014) 574–580

washed with 75% ethanol. The extracted RNA was dissolved in diethylpyrocarbonate-treated water. RNA extraction from skin samples was performed using the RNAeasy Mini Kit (Qiagen, Hilden, Germany) and following the manufacturer’s instruction. For reverse-transcription polymerase chain reaction (rt-PCR), TaqMan Reverse Transcription Reagents (Applied Biosystems, Darmstadt, Germany) were used. The PCR reaction contained 500 ng of mRNA, 10 reaction buffer, deoxynucleotide triphosphate (dNTP), MgCl2, random hexameres, RNAse inhibitor, and MultiScribe Reverse Transcriptase (Applied Biosystems, Darmstadt, Germany) and was run at a volume of 100 lL. The 96-well GeneAmp PCR System 9700 cycler (Applied Biosystems) was used with the following cycler conditions: 10 minutes, 38 °C; 60 minutes, 48 °C; 25 minutes, 95 °C. Quantitative real-time PCR was performed using reagents and cyclers purchased from Applied Biosystems. PCR primers and probes specific for mouse ECE-1 (Assay-ID: Mm01187091_m1), ETAR (Assay-ID: Mm01243722_m1), ETBR (Assay-ID: Mm00432989_m1), and 18s-RNA (endogenous control) were obtained as TaqMan gene expression assays. The reaction contained 12.5 lL TaqMan Universal Master Mix and 1.25 lL of the specific primer in a final volume of 25 lL. The cycler conditions were as follows: 50°, 2 minutes; 95 °C, 10 minutes; 40 cycles, 95 °C, 15 seconds; 60 °C, 1 minute. Each quantitative real-time PCR (qRT-PCR) plate contained a tissue- and primer-specific calibrator sample, which was the sample of an untreated mouse, whose threshold cycles (Ct)-values were next to the calculated

mean of all control samples. The absolute value of the calibrator was set to 1, and all measured samples were related to this sample. Samples were measured as triplicates, except for 18sRNA values, which were tested as duplicates. All plates were analyzed using identical conditions. Data were assessed with the comparative DDCt-method as previously described [47]. 2.5. Statistics Data were analyzed using IBM PASW Statistics 19.0 software (IBM, Ehningen, Germany). Results of behavioral tests are presented as means ± standard deviation; ELISA and qRT-PCR results are presented as median and range. The Kruskal-Wallis test and 2-way repeated-measures ANOVAs followed by Least Significant Difference (LSD) post-hoc test were used to detect differences between groups as appropriate. P < .05 was considered significant. 3. Results 3.1. MP prevents nerve injury-induced mechanical and thermal hypersensitivity and edema in NEP ko mice Three days after CCI, wt mice treated with NS had decreased withdrawal latencies to heat compared to baseline, indicating heat hyperalgesia, as expected (P < .01, Fig. 1A). This difference was no longer present at day 7 after CCI (Fig. 1A). NEP ko mice showed reduced withdrawal latencies to heat at day 3 (P < .001) and

Fig. 1. The line plots illustrate pain behavior and hind paw edema in wild-type (wt) and neprilysin (NEP) knockout (ko) mice. P values given with # symbol represent the comparison of data points with baseline values (A, D, G); P values given with  symbol represent the comparison of data points between treatment groups in wt and NEP ko mice each (B–C, E–F, H–I); n = 5 to 6 per group. (A) Paw withdrawal latencies to heat in wt mice treated with normal saline (NS) were reduced only on day 3 after chronic constriction injury (CCI) compared to baseline (##P < .01). In NEP ko mice treated with NS, paw withdrawal latencies to heat were reduced at day 3 (###P < .001) and at day 7 (###P < .001) after CCI compared to baseline. (B) Paw withdrawal latencies did not differ between wt mice treated with NS or with methylprednisolone (MP). (C) NEP ko mice treated with NS developed reduced paw withdrawal latencies to heat at day 3 (⁄⁄⁄P < .001) and at day 7 (⁄⁄⁄P < .001) after CCI compared to mice treated with MP. MP completely prevented hypersensitivity to heat in NEP ko mice. (D) Paw withdrawal thresholds to mechanical stimulation in wt mice treated with NS were only slightly reduced on day 3 after CCI, but decreased at day 7 compared to baseline (###P < .001). In NEP ko mice treated with NS, paw withdrawal thresholds to mechanical stimuli decreased at day 3 (#P < .05) and were still reduced at day 7 compared to baseline (##P < .01). (E) Paw withdrawal thresholds to mechanical stimulation did not differ between wt mice treated with NS and with MP. (F) NEP ko mice treated with NS showed reduced paw withdrawal thresholds to mechanical stimuli at day 7 (⁄⁄⁄P < .001) after CCI compared to mice treated with MP. MP prevented hypersensitivity to mechanical stimuli in NEP ko mice. (G) Paw thickness did not change significantly in wt mice treated with NS 3 and 7 days after CCI compared to baseline. In NEP ko mice, paw edema was observed at day 3 (###P < .001) and 7 (###P < .001) after CCI compared to baseline. (H) Paw thickness was not different between wt mice treated with NS or with MP. (I) NEP ko mice treated with NS developed paw edema at 3 (⁄⁄⁄P < .001) and 7 (⁄⁄⁄P < .001) days after CCI compared to mice treated with MP. MP protected NEP ko mice from paw edema.

Ò

L. He et al. / PAIN 155 (2014) 574–580

at day 7 after CCI (P < .001, Fig. 1A). In wt mice, treatment with MP did not alter withdrawal latencies compared to NS treatment at days 3 and 7 (Fig. 1B). As shown before, NEP ko mice developed reduced withdrawal latencies to heat 3 (P < .001) and 7 days after CCI when additionally treated with NS (P < .001, Fig. 1C). NEP ko mice treated with MP, however, did not develop hypersensitivity to heat at all after CCI (Fig. 1C). Withdrawal thresholds to mechanical stimulation were only slightly reduced in wt mice 3 days after CCI; however, they were decreased at day 7 compared to baseline (P < .01, Fig. 1D). In NEP ko mice, withdrawal thresholds to mechanical stimuli decreased at day 3 (P < .05) after CCI and were still reduced at day 7 (P < .001) compared to baseline (Fig. 1D). MP treatment did not prevent the reduction of paw withdrawal thresholds in wt mice 3 (P < .05) and 7 days after CCI (P < .001, Fig. 1E), whereas NEP ko mice treated with MP were protected (Fig. 1F). At day 7, mice treated with NS (P < .001) had a much more pronounced reduction in withdrawal thresholds than NEP ko mice treated with MP (P < .05; Fig. 1F). Thus, continuously infused MP protected NEP ko mice from both thermal and mechanical hypersensitivity after CCI, but did not alter postinjury sensitivity in wt mice. Although wt mice treated with NS did not show a significant change in paw thickness 3 and 7 days after CCI compared to baseline, NEP ko mice developed paw edema (P < .001 for each day vs baseline, Fig. 1G). In wt mice, paw thickness also was not different between animals treated with NS or with MP (Fig. 1H). However, in NEP ko mice, continuously infused MP prevented edema (Fig. 1I). 3.2. Neuropeptide content in sciatic nerves Sciatic nerves for neuropeptide analysis were taken from naïve mice and on day 8 after CCI. CGRP protein levels in the sciatic nerve

577

were not different between wt and NEP ko mice, as previously reported [28], and did not change after CCI or MP treatment (Fig. 2A). SP levels were slightly higher in naïve NEP ko mice compared to naïve wt mice (P < .05), and mildly increased only in wt mice after CCI (P < .01). MP did not change SP levels in either genotype (Fig. 2B). Bradykinin levels were higher in naïve NEP ko mice than in naïve wt mice. CCI increased bradykinin levels in NEP ko mice (P < .005), but not in wt mice. MP induced a mild increase of bradykinin in wt mice (P < .05), but no further increase in NEP ko mice (Fig. 2C). ET-1 levels were 3-fold higher in NEP ko mice in the naive state (P < .05) and increased further 2-fold after CCI (P < .05, Fig. 2D). MP did not influence ET-1 levels in either genotype. 3.3. ECE-1 and ET-1 receptor gene expression in sciatic nerve Gene expression of ECE-1 and the ET-1 receptors was measured in naïve wt and NEP ko mice, in mice with MP treatment only, and on day 8 after CCI with MP or saline treatment. ECE-1 gene expression was not different between genotypes, but was increased in both genotypes after CCI with MP treatment and also in naïve mice with MP treatment only (P < .001 each). ECE-1 gene expression did not change in mice after CCI with NS (Fig. 3A). ETAR and ETBR gene expression was not different between genotypes and did not increase after CCI in the ipsilateral sciatic nerve (Fig. 3B and C). In MP-treated mice of both genotypes, ETAR and ETBR gene expression was significantly increased compared to saline-treated CCI mice (P < .001; Fig. 3B and C). This MP-induced increase in ETAR and ETBR was less pronounced in NEP ko mice compared to wt mice (ETAR: P = .03; ETBR: P = .04; Fig. 3A and B). MP treatment of naïve mice produced a significant increase in both genotypes for ETAR and ETBR gene expression (P < .001 each comparing naïve with MP treatment, Fig. 3).

Fig. 2. Protein levels of (A) calcitonin gene-related peptide (CGRP), (B) substance P (SP), (C) bradykinin, (D) endothelin-1 (ET-1) in mouse sciatic nerves, measured by enzymelinked immunosorbent assay. (A) CGRP levels in sciatic nerve are not influenced by chronic constriction injury (CCI) or methylprednisolone (MP) and are not different between wild-type (wt) and neprilysin (NEP) knockout (ko) mice. (B) Substance P protein levels are higher in naïve NEP ko mice than in wt mice. CCI induces substance P protein only in wt mice. Bradykinin levels increased in NEP ko mice after CCI and in both genotypes under MP treatment. ET-1 levels were higher in NEP ko mice in the naive state, after CCI, and after CCI and MP treatment, compared to wt mice. ⁄P < .05; ⁄⁄P < .01.

578

Ò

L. He et al. / PAIN 155 (2014) 574–580

Fig. 4. Bar graphs show the paw withdrawal latency to heat stimulation (A) and the withdrawal threshold to mechanical stimulation (B) in neprilysin (NEP) knockout (ko) mice at baseline and on day 7 after chronic constriction injury. Mice received an intraplantar injection of the ETBR agonist IRL-1620 on the ipsilateral (right) paw, and behavioral tests were repeated 30 minutes after injection. (A) IRL-1620 injection increased paw withdrawal latencies upon heat stimulation in NEP ko mice. (B) Withdrawal thresholds to mechanical stimulation also increased after IRL1620 injection, however, the difference to day 7 without IRL-1620 was not significant. Vehicle injection: 1 day after baseline assessment. ⁄⁄P < .01.

4. Discussion

Fig. 3. Relative gene expression of (A) endothelin-converting enzyme 1 (ECE-1), and the endothelin-1 receptors (B) ETAR and (C) ETBR. Only the combination of chronic constriction injury and methylprednisolone (MP) treatment increased all levels in both genotypes. MP treatment of naïve mice increased ECE-1, ETAR, and ETBR gene expression in both genotypes. MP-induced increase in ETAR and ETBR is less pronounced in neprilysin ko mice compared to wt mice (ETAR: P < .05; ETBR: P < .05; Figs. 3A, 3B). ⁄P < .05, ⁄⁄⁄P < .001.

3.4. The ETBR agonist IRL-1620 reduces pain behavior in NEP ko mice after CCI, whereas the ETBR antagonist partially antagonizes the MP effect Intraplantar application of the ETBR agonist IRL-1620 at day 7 after CCI led to an increase in heat withdrawal latencies in NEP ko mice within 30 minutes (P < .01; Fig. 4A). There also was a marked trend for increased mechanical withdrawal thresholds (not significant; Fig. 4B). In contrast, intraplantar injection of the ETBR antagonist BQ-788 at day 7 after CCI led to a decrease in heat withdrawal latencies in MP-treated NEP ko mice within 30 minutes (not significant; Fig. 5A). There also was a trend for reduced mechanical withdrawal thresholds (not significant; Fig. 5B). The injection of IRL-1620 in the ipsilateral hindpaw of NEP ko mice did not change ETAR and ETBR gene expression in the paw skin within the first 30 minutes (data not shown).

NEP degrades peptides and endorphins that are involved in the regulation of inflammation and pain [8,33]. NEP also is involved in bone remodeling after fracture [22]. We recently showed that NEP inhibition leads to sweating disturbances [40,41]. In CRPS, all of these symptoms coexist and define the clinical presentation. Clinical studies indicate that different peptides may be involved in CRPS pathophysiology and that steroids relieve the clinical symptoms [4,11,23,32]. For a detailed analysis of how all of these findings contribute to CRPS after trauma, animal models are crucial. The existing CRPS models to date [12,26] have severe drawbacks, such as the high prevalence of CRPS-like symptoms after fracture, whereas human posttraumatic CRPS occurs in only 5% of cases. Based on these facts, we set out to investigate an alternative mouse model for CRPS. 4.1. MP prevents hyperalgesia in NEP ko but not wt mice Compared to wt mice, NEP-deficient mice showed increased sensitivity to heat and mechanical stimuli and exaggerated paw edema after CCI, confirming our previous study [28]. Having shown this increased posttraumatic inflammation in NEP ko mice with CCI, we hypothesized that, resembling human CRPS, hyperalgesia in these mice might respond to corticosteroids. We show that continuously infused MP prevented hind paw edema as well as thermal and mechanical hyperalgesia in NEP ko mice after CCI. 4.2. MP reduces edema in NEP ko mice but not in wt mice One reason for the selective antiedematous effect of MP in NEP ko mice may simply be the more prominent edema in these mice,

Ò

L. He et al. / PAIN 155 (2014) 574–580

Fig. 5. Bar graphs show the paw withdrawal latency to heat stimulation (A) and the withdrawal threshold to mechanical stimulation (B) in methylprednisolone -treated neprilysin (NEP) knockout (ko) mice at baseline and on day 7 after chronic constriction injury. Mice received an intraplantar injection of the ETBR antagonist BQ-788 on the ipsilateral (right) paw, and behavioral tests were repeated 30 minutes after injection. (A) BQ-788 injection reduced paw withdrawal latencies upon heat stimulation in NEP ko mice (not significant). (B) Withdrawal thresholds to mechanical stimulation also decreased after BQ-788 injection (not significant). Vehicle injection: 1 day after baseline assessment.

such that a reduction is more easily detectable than in wt mice. On the other hand, there are a number of pathways by which MP may have reduced edema specifically in NEP ko mice. First, glucocorticoids prevent the recruitment of tachykinin receptors [37]. Without steroids, these tachykinin receptors are upregulated on endothelial cells during inflammation [6], thereby amplifying the edematous action of SP, which is released from C-fibers and inflammatory cells [20]. Second, bradykinin is another peptide involved in edema formation [44]. Bradykinin levels are elevated in naïve NEP ko mice and increase further after CCI. There is evidence for a negative interference of glucocorticoids with bradykinin-B2 receptors that is important for the inflammatory bradykinin response [48]. Third, we cannot exclude an MP action entirely unrelated to the currently investigated peptides because MP is also known, for example, to upregulate the vaso pressin-V1 receptor, which mediates vasoconstriction and counteracts edema [13]. 4.3. MP increased ET-1, ETAR, and ETBR Changes in levels of the neuropeptides CGRP, SP, and bradykinin in the sciatic nerve did not provide an explanation for the genotypic differences in MP analgesic effectiveness after CCI. We thus focused our analysis on ET-1, a potent algesic and vasoconstrictive peptide. ET-1 protein levels were higher in nerves from NEP ko than from wt mice in the naïve state and after CCI. This difference between genotypes also was present after MP treatment. Because ET-1 levels are determined by the relative expression of the 2 enzymes ECE-1 and NEP, we investigated the gene expression of ECE-1. We found that MP induces ECE-1 expression as already shown previously [29], but ECE-1 did not differ between genotypes or after CCI. Thus NEP deficiency does not induce an overall increase in ECE-1 expression, as might have been expected as a compensatory mechanism.

579

Because ET-1 actions only can be understood with knowledge about the presence of its receptors, we next measured gene expression of the ET-1 receptors. Both ETAR and ERBR were markedly increased after MP treatment in naïve ko and wt mice as well as in mice with CCI. Animal and human studies indicate that ET-1 may be of importance for the development of pathological pain through binding either solely to one receptor type or to both ETAR and ETBR [19]. ETAR activation seems to always be proalgesic, whereas ETBR can yield proalgesic and analgesic effects depending on tissue concentration. For example, administration of ET-1 to the rat plantar hind paw produces pain-like behavior through activation of ETAR receptors on primary afferent fibers [5,14]. In contrast, high concentrations of local ETBR agonists reduced pain behavior in a chronic postischemic model of CRPS [34]. So far, outside the trigeminal region [1] only ETARs have been shown on neurons and axons, whereas ETBRs were found on glial and endothelial cells [34,38]. The pronociceptive properties of ETAR have been considered to directly result in changes in neuronal ion channels that enhance excitability such as inhibition of delayed rectifier type potassium channels and activation of tetrodotoxin-resistant sodium channels or the transient receptor potential vanilloid type I receptors [9]. The antinociceptive effects of ETBR are indirect: keratinocytes release b-endorphins upon stimulation of ETBR [24,25]. In the context of our present results we speculate that, although ET-1 receptors are present in abundance in both genotypes after MP treatment, the increase in ET-1 in the NEP ko mice acting on the ETBR may be responsible for the analgesic effect of MP. The trend to reduced analgesia in MP-treated NEP ko mice after application of the ETBR antagonist BQ-788, although not directly supporting this speculation, is in favor of it. Furthermore, NEP is one of the enzymes degrading b-endorphin [17]. Thus, it is possible that in NEP ko mice with MP-induced analgesia the elevated ET-1 acts effectively on ETBR, and the b-endorphins subsequently released from keratinocytes are less effectively degraded due to the missing NEP. Previous findings furthermore underscore that low doses of ET-1 preferentially excite ETAR. This sensitizes nociceptors but does not activate the endorphin analgesic pathway [19]; in contrast, higher ET-1 doses activate ETBR and thereby the downstream b-endorphin release [39]. Our finding that intraplantar application of the ETBR agonist IRL-1620 after CCI had an antihyperalgesic effect also supports this argument. We conclude that MP reverses signs of inflammation and hyperalgesia in mice with nerve injury devoid of NEP, a zinc endopeptidase that contributes to the turnover of a plethora of neuropeptides that are postulated to play a key role in CRPS pathophysiology. MP effects are more likely mediated by the regulation of peptide receptors than of the peptides themselves. For the reversal of hyperalgesia, ETBR upregulation might be relevant. Although CCI is not a complete model of CRPS, we suggest that the pathophysiological mechanisms described here may be relevant in acute posttraumatic CRPS with nerve injury. Further studies will focus on the potential importance of these processes in human CRPS. Conflict of interest statement The authors declare the following conflicts of interest. Nurcan Üçeyler: speaker honoraria: Genzyme Corp., Eczacıbasßı-Baxter, Astellas; travel grants: Pfizer Inc., Eczacıbasßı-Baxter, Genzyme Corp., Astellas, Grünenthal GmbH, CSL Behring. Heidrun H. Krämer: speaker honoraria: Pfizer, Actelion. Frank Birklein: speaker honoraria and advisory board membership: Grünenthal, Lilly, Pfizer, Sanofi. Claudia Sommer: speaker honoraria: Astellas, Baxter, CSL Behring, Genzyme Corp., GSK, Pfizer. Advisory boards: Astellas, Genzyme, Pfizer.

580

Ò

L. He et al. / PAIN 155 (2014) 574–580

Acknowledgements The authors thank L. Biko, H. Brünner, B. Dekant, and S. Mildner for expert technical assistance. Funding was provided by research funds of the University of Würzburg. References [1] Barr TP, Kam S, Khodorova A, Montmayeur JP, Strichartz GR. New perspectives on the endothelin axis in pain. Pharmacol Res 2011;63:532–40. [2] Bennett GJ, Xie YK. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. PAINÒ 1988;33:87–107. [3] Blair SJ, Chinthagada M, Hoppenstehdt D, Kijowski R, Fareed J. Role of neuropeptides in pathogenesis of reflex sympathetic dystrophy. Acta Orthop Belg 1998;64:448–51. [4] Braus DF, Krauss JK, Strobel J. The shoulder-hand syndrome after stroke: a prospective clinical trial. Ann Neurol 1994;36:728–33. [5] Carducci MA, Nelson JB, Bowling MK, Rogers T, Eisenberger MA, Sinibaldi V, Donehower R, Leahy TL, Carr RA, Isaacson JD, Janus TJ, Andre A, Hosmane BS, Padley RJ. Atrasentan, an endothelin-receptor antagonist for refractory adenocarcinomas: safety and pharmacokinetics. J Clin Oncol 2002;20: 2171–80. [6] Carlton SM, Coggeshall RE. Inflammation-induced up-regulation of neurokinin 1 receptors in rat glabrous skin. Neurosci Lett 2002;326:29–32. [7] Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL. Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 1994;53:55–63. [8] Chen JJ, Barber LA, Dymshitz J, Vasko MR. Peptidase inhibitors improve recovery of substance P and calcitonin gene-related peptide release from rat spinal cord slices. Peptides 1996;17:31–7. [9] Chichorro JG, Fiuza CR, Bressan E, Claudino RF, Leite DF, Rae GA. Endothelins as pronociceptive mediators of the rat trigeminal system: role of ETA and ETB receptors. Brain Res 2010;1345:73–83. [10] Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156–9. [11] Christensen K, Jensen EM, Noer I. The reflex dystrophy syndrome response to treatment with systemic corticosteroids. Acta Chir Scand 1982;148:653–5. [12] Coderre TJ, Bennett GJ. A hypothesis for the cause of complex regional pain syndrome-type I (reflex sympathetic dystrophy): pain due to deep-tissue microvascular pathology. Pain Med 2010;11:1224–38. [13] Colson P, Ibarondo J, Devilliers G, Balestre MN, Duvoid A, Guillon G. Upregulation of V1a vasopressin receptors by glucocorticoids. Am J Physiol 1992;263:E1054–62. [14] Davar G, Hans G, Fareed MU, Sinnott C, Strichartz G. Behavioral signs of acute pain produced by application of endothelin-1 to rat sciatic nerve. Neuroreport 1998;9:2279–83. [15] Dixon W. The up-and-down method for small samples. J Am Stat Assoc 1965;60:967–78. [16] Dray A, Perkins M. Bradykinin and inflammatory pain. Trends Neurosci 1993; 16:99–104. [17] Gràf L, Paldi A, Patthy A. Action of neutral metalloendopeptidase (‘‘enkephalinase’’) on beta-endorphin. Neuropeptides 1985;6:13–9. [18] Groeneweg JG, Huygen FJ, Heijmans-Antonissen C, Niehof S, Zijlstra FJ. Increased endothelin-1 and diminished nitric oxide levels in blister fluids of patients with intermediate cold type complex regional pain syndrome type 1. BMC Musculoskelet Disord 2006;7:91. [19] Hans G, Schmidt BL, Strichartz G. Nociceptive sensitization by endothelin-1. Brain Res Rev 2009;60:36–42. [20] Holzer P. Neurogenic vasodilatation and plasma leakage in the skin. Gen Pharmacol 1998;30:5–11. [21] Howell S, Caswell AM, Kenny AJ, Turner AJ. Membrane peptidases on human osteoblast-like cells in culture: hydrolysis of calcitonin and hormonal regulation of endopeptidase-24.11. Biochem J 1993;290:159–64. [22] Indig FE, Benayahu D, Fried A, Wientroub S, Blumberg S. Neutral endopeptidase (EC 3.4.24.11) is highly expressed on osteoblastic cells and other marrow stromal cell types. Biochem Biophys Res Commun 1990;172: 620–6. [23] Kalita J, Vajpayee A, Misra UK. Comparison of prednisolone with piroxicam in complex regional pain syndrome following stroke: a randomized controlled trial. QJM 2006;99:89–95. [24] Khodorova A, Navarro B, Jouaville LS, Murphy E, Rice FL, Mazurkiewicz JE, Long-Woodward D, Stoffel M, Strichartz GR, Yukhananov R, Davar G.

[25]

[26]

[27]

[28]

[29]

[30]

[31] [32]

[33]

[34]

[35] [36]

[37] [38]

[39]

[40]

[41]

[42]

[43]

[44] [45]

[46] [47]

[48]

Endothelin-B receptor activation triggers an endogenous analgesic cascade at sites of peripheral injury. Nat Med 2003;9:1055–61. Khodorova A, Richter J, Vasko MR, Strichartz G. Early and late contributions of glutamate and CGRP to mechanical sensitization by endothelin-1. J Pain 2009;10:740–9. Kingery WS, Agashe GS, Sawamura S, Davies MF, Clark JD, Maze M. Glucocorticoid inhibition of neuropathic hyperalgesia and spinal Fos expression. Anesth Analg 2001;92:476–82. Kingery WS, Guo T, Agashe GS, Davies MF, Clark JD, Maze M. Glucocorticoid inhibition of neuropathic limb edema and cutaneous neurogenic extravasation. Brain Res 2001;913:140–8. Krämer HH, He L, Lu B, Birklein F, Sommer C. Increased pain and neurogenic inflammation in mice deficient of neutral endopeptidase. Neurobiol Dis 2009;35:177–83. Lozano E, Segarra M, Corbera-Bellalta M, Garcia-Martinez A, Espigol-Frigole G, Pla-Campo A, Hernandez-Rodriguez J, Cid MC. Increased expression of the endothelin system in arterial lesions from patients with giant-cell arteritis: association between elevated plasma endothelin levels and the development of ischaemic events. Ann Rheum Dis 2010;69:434–42. Lu B, Figini M, Emanueli C, Geppetti P, Grady EF, Gerard NP, Ansell J, Payan DG, Gerard C, Bunnett N. The control of microvascular permeability and blood pressure by neutral endopeptidase. Nat Med 1997;3:904–7. Maihofner C, Seifert F, Markovic K. Complex regional pain syndromes: new pathophysiological concepts and therapies. Eur J Neurol 2010;17:649–60. Marinus J, Moseley GL, Birklein F, Baron R, Maihofner C, Kingery WS, van Hilten JJ. Clinical features and pathophysiology of complex regional pain syndrome. Lancet Neurol 2011;10:637–48. Mentlein R, Roos T. Proteases involved in the metabolism of angiotensin II, bradykinin, calcitonin gene-related peptide (CGRP), and neuropeptide Y by vascular smooth muscle cells. Peptides 1996;17:709–20. Millecamps M, Laferriere A, Ragavendran JV, Stone LS, Coderre TJ. Role of peripheral endothelin receptors in an animal model of complex regional pain syndrome type 1 (CRPS-I). PAINÒ 2010;151:174–83. Naleschinski D, Baron R. Complex regional pain syndrome type I: neuropathic or not? Curr Pain Headache Rep 2010;14:196–202. Nalivaeva NN, Belyaev ND, Zhuravin IA, Turner AJ. The Alzheimer’s amyloiddegrading peptidase, neprilysin: can we control it? Int J Alzheimers Dis 2012;2012:383796. O’Connor TM, O’Connell J, O’Brien DI, Goode T, Bredin CP, Shanahan F. The role of substance P in inflammatory disease. J Cell Physiol 2004;201:167–80. Peters CM, Rogers SD, Pomonis JD, Egnazyck GF, Keyser CP, Schmidt JA, Ghilardi JR, Maggio JE, Mantyh PW. Endothelin receptor expression in the normal and injured spinal cord: potential involvement in injury-induced ischemia and gliosis. Exp Neurol 2003;180:1–13. Roques BP, Fournie-Zaluski MC, Wurm M. Inhibiting the breakdown of endogenous opioids and cannabinoids to alleviate pain. Nat Rev Drug Discov 2012;11:292–310. Ruchon AF, Marcinkiewicz M, Ellefsen K, Basak A, Aubin J, Crine P, Boileau G. Cellular localization of neprilysin in mouse bone tissue and putative role in hydrolysis of osteogenic peptides. J Bone Miner Res 2000;15:1266–74. Schlereth T, Breimhorst M, Werner N, Pottschmidt K, Drummond PD, Birklein F. Inhibition of neuropeptide degradation suppresses sweating but increases the area of the axon reflex flare. Exp Dermatol 2013;22:299–301. Shields SD, Cheng X, Üçeyler N, Sommer C, Dib-Hajj SD, Waxman SG. Sodium channel Na(v)1.7 is essential for lowering heat pain threshold after burn injury. J Neurosci 2012;32:10819–32. Sommer C, Schäfers M. Painful mononeuropathy in C57BL/Wld mice with delayed wallerian degeneration: differential effects of cytokine production and nerve regeneration on thermal and mechanical hypersensitivity. Brain Res 1998;784:154–62. Starr MS, West GB. Bradykinin and oedema formation in heated paws of rats. Br J Pharmacol Chemother 1967;31:178–87. Üçeyler N, Tscharke A, Sommer C. Early cytokine expression in mouse sciatic nerve after chronic constriction nerve injury depends on calpain. Brain Behav Immun 2007;21:553–60. Weber M, Birklein F, Neundorfer B, Schmelz M. Facilitated neurogenic inflammation in complex regional pain syndrome. PAINÒ 2001;91:251–7. Winer J, Jung CK, Shackel I, Williams PM. Development and validation of realtime quantitative reverse transcriptase-polymerase chain reaction for monitoring gene expression in cardiac myocytes in vitro. Anal Biochem 1999;270:41–9. Zhang Y, Adner M, Cardell LO. Glucocorticoids suppress transcriptional upregulation of bradykinin receptors in a murine in vitro model of chronic airway inflammation. Clin Exp Allergy 2005;35:531–8.